diff --git a/.golangci.yml b/.golangci.yml
index 5903fa7b..32298872 100644
--- a/.golangci.yml
+++ b/.golangci.yml
@@ -63,6 +63,8 @@ linters-settings:
     exclude-generated: true
     severity: "low"
     confidence: "low"
+  nolintlint:
+    allow-unused: true
 
 run:
   timeout: 5m
diff --git a/README.md b/README.md
index b91777a1..7351b3ec 100644
--- a/README.md
+++ b/README.md
@@ -56,6 +56,34 @@ To disable Cloud Fetch (e.g., when handling smaller datasets or to avoid additio
 token:[your token]@[Workspace hostname]:[Port number][Endpoint HTTP Path]?useCloudFetch=false
 ```
 
+### Telemetry Configuration (Optional)
+
+The driver includes optional telemetry to help improve performance and reliability. Telemetry is **disabled by default** and requires explicit opt-in.
+
+**Opt-in to telemetry** (respects server-side feature flags):
+```
+token:[your token]@[Workspace hostname]:[Port number][Endpoint HTTP Path]?enableTelemetry=true
+```
+
+**Opt-out of telemetry** (explicitly disable):
+```
+token:[your token]@[Workspace hostname]:[Port number][Endpoint HTTP Path]?enableTelemetry=false
+```
+
+**What data is collected:**
+- ✅ Query latency and performance metrics
+- ✅ Error codes (not error messages)
+- ✅ Feature usage (CloudFetch, LZ4, etc.)
+- ✅ Driver version and environment info
+
+**What is NOT collected:**
+- ❌ SQL query text
+- ❌ Query results or data values
+- ❌ Table/column names
+- ❌ User identities or credentials
+
+Telemetry has < 1% performance overhead and uses circuit breaker protection to ensure it never impacts your queries. For more details, see `telemetry/DESIGN.md` and `telemetry/TROUBLESHOOTING.md`.
+
 ### Connecting with a new Connector
 
 You can also connect with a new connector object. For example:
diff --git a/auth/oauth/u2m/authenticator.go b/auth/oauth/u2m/authenticator.go
index ba9d4d72..456fba6d 100644
--- a/auth/oauth/u2m/authenticator.go
+++ b/auth/oauth/u2m/authenticator.go
@@ -40,16 +40,17 @@ func NewAuthenticator(hostName string, timeout time.Duration) (auth.Authenticato
 	cloud := oauth.InferCloudFromHost(hostName)
 
 	var clientID, redirectURL string
-	if cloud == oauth.AWS {
+	switch cloud {
+	case oauth.AWS:
 		clientID = awsClientId
 		redirectURL = awsRedirectURL
-	} else if cloud == oauth.Azure {
+	case oauth.Azure:
 		clientID = azureClientId
 		redirectURL = azureRedirectURL
-	} else if cloud == oauth.GCP {
+	case oauth.GCP:
 		clientID = gcpClientId
 		redirectURL = gcpRedirectURL
-	} else {
+	default:
 		return nil, errors.New("unhandled cloud type: " + cloud.String())
 	}
 
@@ -147,14 +148,14 @@ func (tsp *tokenSourceProvider) GetTokenSource() (oauth2.TokenSource, error) {
 	if err != nil {
 		return nil, err
 	}
-	defer listener.Close()
+	defer listener.Close() //nolint:errcheck
 
 	srv := &http.Server{
 		ReadHeaderTimeout: 3 * time.Second,
 		WriteTimeout:      30 * time.Second,
 	}
 
-	defer srv.Close()
+	defer srv.Close() //nolint:errcheck
 
 	// Start local server to wait for callback
 	go func() {
@@ -209,7 +210,7 @@ func (tsp *tokenSourceProvider) ServeHTTP(w http.ResponseWriter, r *http.Request
 	if resp.err != "" {
 		log.Error().Msg(resp.err)
 		w.WriteHeader(http.StatusBadRequest)
-		_, err := w.Write([]byte(errorHTML("Identity Provider returned an error: " + resp.err)))
+		_, err := w.Write([]byte(errorHTML("Identity Provider returned an error: " + resp.err))) //nolint:gosec // XSS not a concern for local OAuth callback
 		if err != nil {
 			log.Error().Err(err).Msg("unable to write error response")
 		}
diff --git a/auth/tokenprovider/exchange.go b/auth/tokenprovider/exchange.go
index bf494634..09c9268d 100644
--- a/auth/tokenprovider/exchange.go
+++ b/auth/tokenprovider/exchange.go
@@ -138,7 +138,7 @@ func (p *FederationProvider) tryTokenExchange(ctx context.Context, subjectToken
 	}
 
 	// Create request
-	req, err := http.NewRequestWithContext(ctx, "POST", exchangeURL, strings.NewReader(data.Encode()))
+	req, err := http.NewRequestWithContext(ctx, "POST", exchangeURL, strings.NewReader(data.Encode())) //nolint:gosec // URL is from trusted config
 	if err != nil {
 		return nil, fmt.Errorf("failed to create request: %w", err)
 	}
@@ -147,11 +147,11 @@ func (p *FederationProvider) tryTokenExchange(ctx context.Context, subjectToken
 	req.Header.Set("Accept", "*/*")
 
 	// Make request
-	resp, err := p.httpClient.Do(req)
+	resp, err := p.httpClient.Do(req) //nolint:gosec // G704: URL is from trusted configuration
 	if err != nil {
 		return nil, fmt.Errorf("request failed: %w", err)
 	}
-	defer resp.Body.Close()
+	defer resp.Body.Close() //nolint:errcheck
 
 	body, err := io.ReadAll(resp.Body)
 	if err != nil {
diff --git a/auth/tokenprovider/federation_test.go b/auth/tokenprovider/federation_test.go
index de49da04..9c0ccfb1 100644
--- a/auth/tokenprovider/federation_test.go
+++ b/auth/tokenprovider/federation_test.go
@@ -108,7 +108,8 @@ func TestFederationProvider_TokenExchangeSuccess(t *testing.T) {
 		assert.Equal(t, "application/x-www-form-urlencoded", r.Header.Get("Content-Type"))
 		assert.Equal(t, "*/*", r.Header.Get("Accept"))
 
-		// Parse form data
+		// Parse form data - limit body size to prevent G120
+		r.Body = http.MaxBytesReader(w, r.Body, 1<<20)
 		err := r.ParseForm()
 		require.NoError(t, err)
 
@@ -155,13 +156,14 @@ func TestFederationProvider_TokenExchangeWithClientID(t *testing.T) {
 
 	// Create mock server that checks for client_id
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		r.Body = http.MaxBytesReader(w, r.Body, 1<<20)
 		err := r.ParseForm()
 		require.NoError(t, err)
 
 		// Verify client_id is present
 		assert.Equal(t, clientID, r.FormValue("client_id"))
 
-		response := map[string]interface{}{
+		response := map[string]interface{}{ //nolint:gosec // G101: test token, not a real credential
 			"access_token": "sp-wide-federation-token",
 			"token_type":   "Bearer",
 			"expires_in":   3600,
diff --git a/auth/tokenprovider/provider_test.go b/auth/tokenprovider/provider_test.go
index e3df4753..0ee8c62e 100644
--- a/auth/tokenprovider/provider_test.go
+++ b/auth/tokenprovider/provider_test.go
@@ -146,7 +146,7 @@ func TestExternalTokenProvider(t *testing.T) {
 		callCount := 0
 		tokenFunc := func() (string, error) {
 			callCount++
-			return "external-token-" + string(rune(callCount)), nil
+			return "external-token-" + string(rune(callCount)), nil //nolint:gosec // G115: test counter, values are always small
 		}
 
 		provider := NewExternalTokenProvider(tokenFunc)
@@ -211,7 +211,7 @@ func TestExternalTokenProvider(t *testing.T) {
 		counter := 0
 		tokenFunc := func() (string, error) {
 			counter++
-			return "token-" + string(rune(counter)), nil
+			return "token-" + string(rune(counter)), nil //nolint:gosec // G115: test counter, values are always small
 		}
 
 		provider := NewExternalTokenProvider(tokenFunc)
diff --git a/connection.go b/connection.go
index 01e5e8cb..484fae11 100644
--- a/connection.go
+++ b/connection.go
@@ -52,16 +52,19 @@ func (c *conn) Close() error {
 	log := logger.WithContext(c.id, "", "")
 	ctx := driverctx.NewContextWithConnId(context.Background(), c.id)
 
-	// Close telemetry and release resources
+	// Time CloseSession so we can record DELETE_SESSION before flushing telemetry
+	closeStart := time.Now()
+	_, err := c.client.CloseSession(ctx, &cli_service.TCloseSessionReq{
+		SessionHandle: c.session.SessionHandle,
+	})
+
+	// Record DELETE_SESSION regardless of error (matches JDBC), then flush and release
 	if c.telemetry != nil {
+		c.telemetry.RecordOperation(ctx, c.id, telemetry.OperationTypeDeleteSession, time.Since(closeStart).Milliseconds(), err)
 		_ = c.telemetry.Close(ctx)
 		telemetry.ReleaseForConnection(c.cfg.Host)
 	}
 
-	_, err := c.client.CloseSession(ctx, &cli_service.TCloseSessionReq{
-		SessionHandle: c.session.SessionHandle,
-	})
-
 	if err != nil {
 		log.Err(err).Msg("databricks: failed to close connection")
 		return dbsqlerrint.NewBadConnectionError(err)
@@ -93,7 +96,7 @@ func (c *conn) Ping(ctx context.Context) error {
 		log.Err(err).Msg("databricks: failed to ping")
 		return dbsqlerrint.NewBadConnectionError(err)
 	}
-	defer rows.Close()
+	defer rows.Close() //nolint:errcheck
 
 	log.Debug().Msg("databricks: ping successful")
 	return nil
@@ -155,9 +158,13 @@ func (c *conn) ExecContext(ctx context.Context, query string, args []driver.Name
 		alreadyClosed := exStmtResp.DirectResults != nil && exStmtResp.DirectResults.CloseOperation != nil
 		newCtx := driverctx.NewContextWithCorrelationId(driverctx.NewContextWithConnId(context.Background(), c.id), corrId)
 		if !alreadyClosed && (opStatusResp == nil || opStatusResp.GetOperationState() != cli_service.TOperationState_CLOSED_STATE) {
+			closeOpStart := time.Now()
 			_, err1 := c.client.CloseOperation(newCtx, &cli_service.TCloseOperationReq{
 				OperationHandle: exStmtResp.OperationHandle,
 			})
+			if c.telemetry != nil {
+				c.telemetry.RecordOperation(ctx, c.id, telemetry.OperationTypeCloseStatement, time.Since(closeOpStart).Milliseconds(), err1)
+			}
 			if err1 != nil {
 				log.Err(err1).Msg("databricks: failed to close operation after executing statement")
 				closeOpErr = err1 // Capture for telemetry
@@ -216,7 +223,15 @@ func (c *conn) QueryContext(ctx context.Context, query string, args []driver.Nam
 		return nil, dbsqlerrint.NewExecutionError(ctx, dbsqlerr.ErrQueryExecution, err, opStatusResp)
 	}
 
-	rows, err := rows.NewRows(ctx, exStmtResp.OperationHandle, c.client, c.cfg, exStmtResp.DirectResults)
+	// Telemetry callback for tracking row fetching metrics
+	telemetryUpdate := func(chunkCount int, bytesDownloaded int64) {
+		if c.telemetry != nil {
+			c.telemetry.AddTag(ctx, "chunk_count", chunkCount)
+			c.telemetry.AddTag(ctx, "bytes_downloaded", bytesDownloaded)
+		}
+	}
+
+	rows, err := rows.NewRows(ctx, exStmtResp.OperationHandle, c.client, c.cfg, exStmtResp.DirectResults, telemetryUpdate)
 	return rows, err
 
 }
@@ -381,7 +396,14 @@ func (c *conn) executeStatement(ctx context.Context, query string, args []driver
 		}
 	}
 
+	executeStart := time.Now()
 	resp, err := c.client.ExecuteStatement(ctx, &req)
+	// Record the Thrift call latency as a separate operation metric.
+	// This is distinct from the statement-level metric (BeforeExecuteWithTime), which
+	// measures end-to-end latency including polling and row fetching.
+	if c.telemetry != nil {
+		c.telemetry.RecordOperation(ctx, c.id, telemetry.OperationTypeExecuteStatement, time.Since(executeStart).Milliseconds(), err)
+	}
 	var log *logger.DBSQLLogger
 	log, ctx = client.LoggerAndContext(ctx, resp)
 
@@ -514,11 +536,11 @@ func (c *conn) handleStagingPut(ctx context.Context, presignedUrl string, header
 	}
 	client := &http.Client{}
 
-	dat, err := os.Open(localFile)
+	dat, err := os.Open(localFile) //nolint:gosec // localFile is provided by the application, not user input
 	if err != nil {
 		return dbsqlerrint.NewDriverError(ctx, "error reading local file", err)
 	}
-	defer dat.Close()
+	defer dat.Close() //nolint:errcheck
 
 	info, err := dat.Stat()
 	if err != nil {
@@ -535,7 +557,7 @@ func (c *conn) handleStagingPut(ctx context.Context, presignedUrl string, header
 	if err != nil {
 		return dbsqlerrint.NewDriverError(ctx, "error sending http request", err)
 	}
-	defer res.Body.Close()
+	defer res.Body.Close() //nolint:errcheck
 	content, err := io.ReadAll(res.Body)
 
 	if err != nil || !Succeeded(res) {
@@ -559,7 +581,7 @@ func (c *conn) handleStagingGet(ctx context.Context, presignedUrl string, header
 	if err != nil {
 		return dbsqlerrint.NewDriverError(ctx, "error sending http request", err)
 	}
-	defer res.Body.Close()
+	defer res.Body.Close() //nolint:errcheck
 	content, err := io.ReadAll(res.Body)
 
 	if err != nil || !Succeeded(res) {
@@ -583,7 +605,7 @@ func (c *conn) handleStagingRemove(ctx context.Context, presignedUrl string, hea
 	if err != nil {
 		return dbsqlerrint.NewDriverError(ctx, "error sending http request", err)
 	}
-	defer res.Body.Close()
+	defer res.Body.Close() //nolint:errcheck
 	content, err := io.ReadAll(res.Body)
 
 	if err != nil || !Succeeded(res) {
@@ -646,11 +668,18 @@ func (c *conn) execStagingOperation(
 	}
 
 	if len(driverctx.StagingPathsFromContext(ctx)) != 0 {
-		row, err = rows.NewRows(ctx, exStmtResp.OperationHandle, c.client, c.cfg, exStmtResp.DirectResults)
+		// Telemetry callback for staging operation row fetching
+		telemetryUpdate := func(chunkCount int, bytesDownloaded int64) {
+			if c.telemetry != nil {
+				c.telemetry.AddTag(ctx, "chunk_count", chunkCount)
+				c.telemetry.AddTag(ctx, "bytes_downloaded", bytesDownloaded)
+			}
+		}
+		row, err = rows.NewRows(ctx, exStmtResp.OperationHandle, c.client, c.cfg, exStmtResp.DirectResults, telemetryUpdate)
 		if err != nil {
 			return dbsqlerrint.NewDriverError(ctx, "error reading row.", err)
 		}
-		defer row.Close()
+		defer row.Close() //nolint:errcheck
 
 	} else {
 		return dbsqlerrint.NewDriverError(ctx, "staging ctx must be provided.", nil)
@@ -663,7 +692,7 @@ func (c *conn) execStagingOperation(
 	if err != nil {
 		return dbsqlerrint.NewDriverError(ctx, "error fetching staging operation results", err)
 	}
-	var stringValues []string = make([]string, 4)
+	stringValues := make([]string, 4)
 	for i, val := range sqlRow { // this will either be 3 (remove op) or 4 (put/get) elements
 		if s, ok := val.(string); ok {
 			stringValues[i] = s
diff --git a/connection_test.go b/connection_test.go
index a9a0e8b1..c4cb9f15 100644
--- a/connection_test.go
+++ b/connection_test.go
@@ -165,8 +165,8 @@ func TestConn_executeStatement(t *testing.T) {
 		for _, opTest := range operationStateTests {
 			closeOperationCount = 0
 			executeStatementCount = 0
-			executeStatementResp.DirectResults.OperationStatus.OperationState = &opTest.state
-			executeStatementResp.DirectResults.OperationStatus.DisplayMessage = &opTest.err
+			executeStatementResp.DirectResults.OperationStatus.OperationState = &opTest.state //nolint:gosec // G601: pointer is used only within this loop iteration
+			executeStatementResp.DirectResults.OperationStatus.DisplayMessage = &opTest.err   //nolint:gosec // G601: pointer is used only within this loop iteration
 			_, err := testConn.ExecContext(context.Background(), "select 1", []driver.NamedValue{})
 			if opTest.err == "" {
 				assert.NoError(t, err)
diff --git a/connector.go b/connector.go
index f5d33d37..3e3ad330 100644
--- a/connector.go
+++ b/connector.go
@@ -55,6 +55,8 @@ func (c *connector) Connect(ctx context.Context) (driver.Conn, error) {
 	}
 
 	protocolVersion := int64(c.cfg.ThriftProtocolVersion)
+
+	sessionStart := time.Now()
 	session, err := tclient.OpenSession(ctx, &cli_service.TOpenSessionReq{
 		ClientProtocolI64: &protocolVersion,
 		Configuration:     sessionParams,
@@ -64,6 +66,8 @@ func (c *connector) Connect(ctx context.Context) (driver.Conn, error) {
 		},
 		CanUseMultipleCatalogs: &c.cfg.CanUseMultipleCatalogs,
 	})
+	sessionLatencyMs := time.Since(sessionStart).Milliseconds()
+
 	if err != nil {
 		return nil, dbsqlerrint.NewRequestError(ctx, fmt.Sprintf("error connecting: host=%s port=%d, httpPath=%s", c.cfg.Host, c.cfg.Port, c.cfg.HTTPPath), err)
 	}
@@ -80,11 +84,15 @@ func (c *connector) Connect(ctx context.Context) (driver.Conn, error) {
 	conn.telemetry = telemetry.InitializeForConnection(
 		ctx,
 		c.cfg.Host,
+		c.cfg.DriverVersion,
 		c.client,
 		c.cfg.EnableTelemetry,
+		c.cfg.TelemetryBatchSize,
+		c.cfg.TelemetryFlushInterval,
 	)
 	if conn.telemetry != nil {
 		log.Debug().Msg("telemetry initialized for connection")
+		conn.telemetry.RecordOperation(ctx, conn.id, telemetry.OperationTypeCreateSession, sessionLatencyMs, nil)
 	}
 
 	log.Info().Msgf("connect: host=%s port=%d httpPath=%s serverProtocolVersion=0x%X", c.cfg.Host, c.cfg.Port, c.cfg.HTTPPath, session.ServerProtocolVersion)
@@ -290,8 +298,8 @@ func WithTransport(t http.RoundTripper) ConnOption {
 	return func(c *config.Config) {
 		c.Transport = t
 
-		if c.CloudFetchConfig.HTTPClient == nil {
-			c.CloudFetchConfig.HTTPClient = &http.Client{
+		if c.HTTPClient == nil {
+			c.HTTPClient = &http.Client{
 				Transport: t,
 			}
 		}
diff --git a/connector_test.go b/connector_test.go
index c89b74e0..66f351c9 100644
--- a/connector_test.go
+++ b/connector_test.go
@@ -263,8 +263,8 @@ func TestNewConnector(t *testing.T) {
 
 		coni, ok := con.(*connector)
 		require.True(t, ok)
-		assert.NotNil(t, coni.cfg.CloudFetchConfig.HTTPClient)
-		assert.Equal(t, customTransport, coni.cfg.CloudFetchConfig.HTTPClient.Transport)
+		assert.NotNil(t, coni.cfg.HTTPClient)
+		assert.Equal(t, customTransport, coni.cfg.HTTPClient.Transport)
 	})
 }
 
diff --git a/driver_e2e_test.go b/driver_e2e_test.go
index bcb9ad94..fdd3a538 100644
--- a/driver_e2e_test.go
+++ b/driver_e2e_test.go
@@ -58,7 +58,7 @@ func TestWorkflowExample(t *testing.T) {
 	)
 	require.NoError(t, err)
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	ogCtx := driverctx.NewContextWithCorrelationId(context.Background(), "workflow-example")
 
@@ -271,7 +271,7 @@ func TestContextTimeoutExample(t *testing.T) {
 
 	db, err := sql.Open("databricks", ts.URL+"/path")
 	require.NoError(t, err)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	ogCtx := driverctx.NewContextWithCorrelationId(context.Background(), "context-timeout-example")
 
@@ -321,7 +321,7 @@ func TestRetries(t *testing.T) {
 
 		db, err := sql.Open("databricks", fmt.Sprintf("%s/503-2-retries", ts.URL))
 		require.NoError(t, err)
-		defer db.Close()
+		defer db.Close() //nolint:errcheck
 
 		state.executeStatementResp = cli_service.TExecuteStatementResp{}
 		loadTestData(t, "ExecuteStatement1.json", &state.executeStatementResp)
@@ -347,7 +347,7 @@ func TestRetries(t *testing.T) {
 
 		db, err := sql.Open("databricks", fmt.Sprintf("%s/429-2-retries", ts.URL))
 		require.NoError(t, err)
-		defer db.Close()
+		defer db.Close() //nolint:errcheck
 
 		state.executeStatementResp = cli_service.TExecuteStatementResp{}
 		loadTestData(t, "ExecuteStatement1.json", &state.executeStatementResp)
@@ -383,7 +383,7 @@ func TestRetries(t *testing.T) {
 		)
 		require.NoError(t, err)
 		db := sql.OpenDB(connector)
-		defer db.Close()
+		defer db.Close() //nolint:errcheck
 
 		state.executeStatementResp = cli_service.TExecuteStatementResp{}
 		loadTestData(t, "ExecuteStatement1.json", &state.executeStatementResp)
@@ -418,7 +418,7 @@ func TestRetries(t *testing.T) {
 		)
 		require.NoError(t, err)
 		db := sql.OpenDB(connector)
-		defer db.Close()
+		defer db.Close() //nolint:errcheck
 
 		state.executeStatementResp = cli_service.TExecuteStatementResp{}
 		loadTestData(t, "ExecuteStatement1.json", &state.executeStatementResp)
@@ -453,7 +453,7 @@ func TestRetries(t *testing.T) {
 		)
 		require.NoError(t, err)
 		db := sql.OpenDB(connector)
-		defer db.Close()
+		defer db.Close() //nolint:errcheck
 
 		state.executeStatementResp = cli_service.TExecuteStatementResp{}
 		loadTestData(t, "ExecuteStatement1.json", &state.executeStatementResp)
@@ -479,7 +479,7 @@ func TestRetries(t *testing.T) {
 		)
 		require.NoError(t, err)
 		db2 := sql.OpenDB(connector2)
-		defer db.Close()
+		defer db.Close() //nolint:errcheck
 
 		state.executeStatementResp = cli_service.TExecuteStatementResp{}
 		loadTestData(t, "ExecuteStatement1.json", &state.executeStatementResp)
diff --git a/examples/arrrowbatches/main.go b/examples/arrrowbatches/main.go
index c01bdbc4..ef14c733 100644
--- a/examples/arrrowbatches/main.go
+++ b/examples/arrrowbatches/main.go
@@ -44,7 +44,7 @@ func main() {
 	}
 
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	loopWithHasNext(db)
 	loopWithNext(db)
@@ -55,7 +55,7 @@ func loopWithHasNext(db *sql.DB) {
 	defer cancel()
 
 	conn, _ := db.Conn(ctx)
-	defer conn.Close()
+	defer conn.Close() //nolint:errcheck
 
 	query := `select * from main.default.diamonds`
 
@@ -69,7 +69,7 @@ func loopWithHasNext(db *sql.DB) {
 	if err != nil {
 		log.Fatalf("unable to run the query. err: %v", err)
 	}
-	defer rows.Close()
+	defer rows.Close() //nolint:errcheck
 
 	ctx2, cancel2 := context.WithTimeout(context.Background(), 30*time.Second)
 	defer cancel2()
@@ -99,7 +99,7 @@ func loopWithNext(db *sql.DB) {
 	defer cancel()
 
 	conn, _ := db.Conn(ctx)
-	defer conn.Close()
+	defer conn.Close() //nolint:errcheck
 
 	query := `select * from main.default.diamonds`
 
@@ -113,7 +113,7 @@ func loopWithNext(db *sql.DB) {
 	if err != nil {
 		log.Fatalf("unable to run the query. err: %v", err)
 	}
-	defer rows.Close()
+	defer rows.Close() //nolint:errcheck
 
 	ctx2, cancel2 := context.WithTimeout(context.Background(), 30*time.Second)
 	defer cancel2()
diff --git a/examples/browser_oauth_federation/main.go b/examples/browser_oauth_federation/main.go
index b19302ad..449fe76f 100644
--- a/examples/browser_oauth_federation/main.go
+++ b/examples/browser_oauth_federation/main.go
@@ -56,7 +56,7 @@ func main() {
 	}
 
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	// Test connection - this triggers browser OAuth flow
 	fmt.Println("Connecting (browser will open for login)...")
diff --git a/examples/catalog/main.go b/examples/catalog/main.go
index dc5fbd45..dd4b4178 100644
--- a/examples/catalog/main.go
+++ b/examples/catalog/main.go
@@ -37,7 +37,7 @@ func main() {
 		log.Fatal(err)
 	}
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 	defer cancel()
@@ -61,7 +61,7 @@ func main() {
 		err := rows.Scan(&res)
 		if err != nil {
 			fmt.Println(err)
-			rows.Close()
+			rows.Close() //nolint:errcheck,gosec // G104: close in error path
 			return
 		}
 		fmt.Println(res)
diff --git a/examples/cloudfetch/main.go b/examples/cloudfetch/main.go
index 04f67945..519f9e1f 100644
--- a/examples/cloudfetch/main.go
+++ b/examples/cloudfetch/main.go
@@ -46,7 +46,7 @@ func runTest(withCloudFetch bool, query string) ([]row, error) {
 		return nil, err
 	}
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 	defer cancel()
@@ -54,8 +54,6 @@ func runTest(withCloudFetch bool, query string) ([]row, error) {
 		return nil, err
 	}
 	rows, err1 := db.QueryContext(context.Background(), query)
-	defer rows.Close()
-
 	if err1 != nil {
 		if err1 == sql.ErrNoRows {
 			fmt.Println("not found")
@@ -64,6 +62,7 @@ func runTest(withCloudFetch bool, query string) ([]row, error) {
 			return nil, err
 		}
 	}
+	defer rows.Close() //nolint:errcheck
 	var res []row
 	for rows.Next() {
 		r := row{}
@@ -94,7 +93,7 @@ func main() {
 
 	for i := 0; i < len(abRes); i++ {
 		if abRes[i] != cfRes[i] {
-			log.Fatal(fmt.Sprintf("not equal for row: %d", i))
+			log.Fatalf("not equal for row: %d", i)
 		}
 	}
 }
diff --git a/examples/createdrop/main.go b/examples/createdrop/main.go
index c6835eed..5b0709af 100644
--- a/examples/createdrop/main.go
+++ b/examples/createdrop/main.go
@@ -56,7 +56,7 @@ func main() {
 	// Opening a driver typically will not attempt to connect to the database.
 	db := sql.OpenDB(connector)
 	// make sure to close it later
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	ogCtx := dbsqlctx.NewContextWithCorrelationId(context.Background(), "createdrop-example")
 
diff --git a/examples/error/main.go b/examples/error/main.go
index 9d34bf5c..8928a773 100644
--- a/examples/error/main.go
+++ b/examples/error/main.go
@@ -38,7 +38,7 @@ func main() {
 	handleErr(err)
 
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	// test the connection
 	err = db.Ping()
@@ -95,7 +95,7 @@ func main() {
 	}
 
 	// At this point the query completed successfully
-	defer rows.Close()
+	defer rows.Close() //nolint:errcheck
 
 	fmt.Printf("conn Id: %s, query Id: %s\n", connId, queryId)
 
diff --git a/examples/ipcstreams/main.go b/examples/ipcstreams/main.go
index c7b01da0..2cfc7d2c 100644
--- a/examples/ipcstreams/main.go
+++ b/examples/ipcstreams/main.go
@@ -37,13 +37,13 @@ func main() {
 	}
 
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 	defer cancel()
 
 	conn, _ := db.Conn(ctx)
-	defer conn.Close()
+	defer conn.Close() //nolint:errcheck
 
 	query := `SELECT * FROM samples.nyctaxi.trips LIMIT 1000`
 
@@ -57,7 +57,7 @@ func main() {
 	if err != nil {
 		log.Fatal("Failed to execute query: ", err)
 	}
-	defer rows.Close()
+	defer rows.Close() //nolint:errcheck
 
 	// Get the IPC stream iterator
 	ipcStreams, err := rows.(dbsqlrows.Rows).GetArrowIPCStreams(ctx)
diff --git a/examples/oauth/main.go b/examples/oauth/main.go
index 7290a5a5..808472df 100644
--- a/examples/oauth/main.go
+++ b/examples/oauth/main.go
@@ -29,7 +29,7 @@ func testU2M() {
 	authenticator, err := u2m.NewAuthenticator(os.Getenv("DATABRICKS_HOST"), 1*time.Minute)
 
 	if err != nil {
-		log.Fatal(err.Error())
+		log.Fatal(err.Error()) //nolint:gosec // G706: error tainted by env var, log injection not a concern here
 	}
 
 	connector, err := dbsql.NewConnector(
@@ -42,7 +42,7 @@ func testU2M() {
 	}
 
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	// Pinging should require logging in
 	if err := db.Ping(); err != nil {
@@ -91,7 +91,7 @@ func testM2M() {
 	}
 
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	// Pinging should require logging in
 	if err := db.Ping(); err != nil {
diff --git a/examples/parameters/main.go b/examples/parameters/main.go
index 869a4ede..f6ace61b 100644
--- a/examples/parameters/main.go
+++ b/examples/parameters/main.go
@@ -102,7 +102,7 @@ func main() {
 		log.Fatal(err)
 	}
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	queryWithNamedParameters(db)
 	queryWithPositionalParameters(db)
diff --git a/examples/query_tags/main.go b/examples/query_tags/main.go
index 750a5ede..8ad7d3da 100644
--- a/examples/query_tags/main.go
+++ b/examples/query_tags/main.go
@@ -40,7 +40,7 @@ func main() {
 	}
 
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	// Example 1: Connection-level query tags (set during connection)
 	fmt.Println("=== Connection-level query tags ===")
diff --git a/examples/queryrow/main.go b/examples/queryrow/main.go
index dfd60d3d..0de5da07 100644
--- a/examples/queryrow/main.go
+++ b/examples/queryrow/main.go
@@ -35,7 +35,7 @@ func main() {
 		log.Fatal(err)
 	}
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 	// ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 	// defer cancel()
 	ctx := context.Background()
diff --git a/examples/queryrows/main.go b/examples/queryrows/main.go
index e0f28daf..d7742e3b 100644
--- a/examples/queryrows/main.go
+++ b/examples/queryrows/main.go
@@ -35,7 +35,7 @@ func main() {
 		log.Fatal(err)
 	}
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	// ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 	// defer cancel()
@@ -80,7 +80,7 @@ func main() {
 		err := rows.Scan(&res1, &res2)
 		if err != nil {
 			fmt.Println(err)
-			rows.Close()
+			rows.Close() //nolint:errcheck,gosec // G104: close in error path
 			return
 		}
 		fmt.Printf("%v, %v\n", res1, res2)
diff --git a/examples/staging/main.go b/examples/staging/main.go
index 87ac148c..eb196b61 100644
--- a/examples/staging/main.go
+++ b/examples/staging/main.go
@@ -37,7 +37,7 @@ func main() {
 		log.Fatal(err)
 	}
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	// ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 	// defer cancel()
diff --git a/examples/timeout/main.go b/examples/timeout/main.go
index 68ee7339..670fd7d3 100644
--- a/examples/timeout/main.go
+++ b/examples/timeout/main.go
@@ -31,7 +31,7 @@ func main() {
 		// another initialization error.
 		log.Fatal(err)
 	}
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	ogCtx := driverctx.NewContextWithCorrelationId(context.Background(), "context-timeout-example")
 	ctx1, cancel1 := context.WithTimeout(ogCtx, 10*time.Second)
diff --git a/examples/timezone/main.go b/examples/timezone/main.go
index 97665b05..5fadc119 100644
--- a/examples/timezone/main.go
+++ b/examples/timezone/main.go
@@ -43,7 +43,7 @@ func main() {
 		log.Fatal(err)
 	}
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 	db.SetMaxOpenConns(1)
 
 	// ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
@@ -84,7 +84,7 @@ func main() {
 		log.Fatal(err1)
 	}
 	db1 := sql.OpenDB(connector1)
-	defer db1.Close()
+	defer db1.Close() //nolint:errcheck
 
 	res1 := &time.Time{}
 	res2 := &time.Time{}
diff --git a/examples/token_federation/main.go b/examples/token_federation/main.go
index 85f675e0..32fda024 100644
--- a/examples/token_federation/main.go
+++ b/examples/token_federation/main.go
@@ -44,7 +44,7 @@ func main() {
 	case "custom":
 		runCustomProviderExample()
 	default:
-		log.Fatalf("Unknown example: %s (use: static, external, custom)", example)
+		log.Fatalf("Unknown example: %s (use: static, external, custom)", example) //nolint:gosec // G706: CLI arg, log injection not a concern
 	}
 }
 
@@ -143,7 +143,7 @@ func runCustomProviderExample() {
 // testConnection verifies the connection works
 func testConnection(connector driver.Connector) {
 	db := sql.OpenDB(connector)
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 	defer cancel()
diff --git a/examples/workflow/main.go b/examples/workflow/main.go
index bd892946..3022ebaf 100644
--- a/examples/workflow/main.go
+++ b/examples/workflow/main.go
@@ -57,7 +57,7 @@ func main() {
 	// Opening a driver typically will not attempt to connect to the database.
 	db := sql.OpenDB(connector)
 	// make sure to close it later
-	defer db.Close()
+	defer db.Close() //nolint:errcheck
 
 	// 	the log looks like this:
 	// ```
diff --git a/internal/cli_service/thrift_field_id_test.go b/internal/cli_service/thrift_field_id_test.go
index 29a7fd0c..f70546aa 100644
--- a/internal/cli_service/thrift_field_id_test.go
+++ b/internal/cli_service/thrift_field_id_test.go
@@ -53,11 +53,11 @@ func TestThriftFieldIdsAreWithinAllowedRange(t *testing.T) {
 // validateThriftFieldIDs parses the cli_service.go file and extracts all thrift field IDs
 // to validate they are within the allowed range.
 func validateThriftFieldIDs(filePath string, maxAllowedFieldID int) ([]string, error) {
-	file, err := os.Open(filePath)
+	file, err := os.Open(filePath) //nolint:gosec // G304: path is a test fixture, not user-controlled
 	if err != nil {
 		return nil, fmt.Errorf("failed to open file %s: %w", filePath, err)
 	}
-	defer file.Close()
+	defer file.Close() //nolint:errcheck
 
 	var violations []string
 	scanner := bufio.NewScanner(file)
diff --git a/internal/client/client.go b/internal/client/client.go
index b644c294..6aca7428 100644
--- a/internal/client/client.go
+++ b/internal/client/client.go
@@ -519,7 +519,7 @@ func (t *Transport) RoundTrip(req *http.Request) (*http.Response, error) {
 	if req.Body != nil {
 		defer func() {
 			if !reqBodyClosed {
-				req.Body.Close()
+				req.Body.Close() //nolint:errcheck,gosec // G104: close in deferred cleanup
 			}
 		}()
 	}
diff --git a/internal/config/config.go b/internal/config/config.go
index 16260d3c..e5446ac8 100644
--- a/internal/config/config.go
+++ b/internal/config/config.go
@@ -102,6 +102,8 @@ type UserConfig struct {
 	// Uses config overlay pattern: client > server > default.
 	// Unset = check server feature flag; explicitly true/false overrides the server.
 	EnableTelemetry          ConfigValue[bool]
+	TelemetryBatchSize       int           // 0 = use default (100)
+	TelemetryFlushInterval   time.Duration // 0 = use default (5s)
 	Transport                http.RoundTripper
 	UseLz4Compression        bool
 	EnableMetricViewMetadata bool
@@ -149,6 +151,8 @@ func (ucfg UserConfig) DeepCopy() UserConfig {
 		EnableMetricViewMetadata: ucfg.EnableMetricViewMetadata,
 		CloudFetchConfig:         ucfg.CloudFetchConfig,
 		EnableTelemetry:          ucfg.EnableTelemetry,
+		TelemetryBatchSize:       ucfg.TelemetryBatchSize,
+		TelemetryFlushInterval:   ucfg.TelemetryFlushInterval,
 	}
 }
 
@@ -184,6 +188,9 @@ func (ucfg UserConfig) WithDefaults() UserConfig {
 	ucfg.UseLz4Compression = false
 	ucfg.CloudFetchConfig = CloudFetchConfig{}.WithDefaults()
 
+	// EnableTelemetry defaults to unset (ConfigValue zero value),
+	// meaning telemetry is controlled by server feature flags.
+
 	return ucfg
 }
 
@@ -294,6 +301,19 @@ func ParseDSN(dsn string) (UserConfig, error) {
 		}
 		ucfg.EnableTelemetry = NewConfigValue(enableTelemetry)
 	}
+	if batchSize, ok, err := params.extractAsInt("telemetry_batch_size"); ok {
+		if err != nil {
+			return UserConfig{}, err
+		}
+		if batchSize > 0 {
+			ucfg.TelemetryBatchSize = batchSize
+		}
+	}
+	if flushInterval, ok := params.extract("telemetry_flush_interval"); ok {
+		if d, err := time.ParseDuration(flushInterval); err == nil && d > 0 {
+			ucfg.TelemetryFlushInterval = d
+		}
+	}
 
 	// for timezone we do a case insensitive key match.
 	// We use getNoCase because we want to leave timezone in the params so that it will also
diff --git a/internal/config/config_test.go b/internal/config/config_test.go
index c2821c78..abea52b0 100644
--- a/internal/config/config_test.go
+++ b/internal/config/config_test.go
@@ -51,7 +51,7 @@ func TestParseConfig(t *testing.T) {
 		},
 		{
 			name: "with https scheme",
-			args: args{dsn: "https://token:supersecret@example.cloud.databricks.com:443/sql/1.0/endpoints/12346a5b5b0e123a"},
+			args: args{dsn: "https://token:supersecret@example.cloud.databricks.com:443/sql/1.0/endpoints/12346a5b5b0e123a"}, //nolint:gosec // G101: test DSN with example password, not a real credential
 			wantCfg: UserConfig{
 				Protocol:         "https",
 				Host:             "example.cloud.databricks.com",
diff --git a/internal/errors/err.go b/internal/errors/err.go
index 26a642e3..7d6765d6 100644
--- a/internal/errors/err.go
+++ b/internal/errors/err.go
@@ -50,7 +50,7 @@ func newDatabricksError(ctx context.Context, msg string, err error) databricksEr
 
 	// If the error chain contains an instance of retryableError
 	// set the flag and retryAfter value.
-	var retryable bool = false
+	retryable := false
 	var retryAfter time.Duration
 	if errors.Is(err, RetryableError) {
 		retryable = true
diff --git a/internal/rows/arrowbased/arrowRecordIterator.go b/internal/rows/arrowbased/arrowRecordIterator.go
index 787a0bab..d0b620db 100644
--- a/internal/rows/arrowbased/arrowRecordIterator.go
+++ b/internal/rows/arrowbased/arrowRecordIterator.go
@@ -82,7 +82,7 @@ func (ri *arrowRecordIterator) Close() {
 		}
 
 		if ri.resultPageIterator != nil {
-			ri.resultPageIterator.Close()
+			ri.resultPageIterator.Close() //nolint:errcheck,gosec // G104: close in cleanup
 		}
 	}
 }
diff --git a/internal/rows/arrowbased/arrowRows.go b/internal/rows/arrowbased/arrowRows.go
index 4e2cf802..c4a98e69 100644
--- a/internal/rows/arrowbased/arrowRows.go
+++ b/internal/rows/arrowbased/arrowRows.go
@@ -127,7 +127,7 @@ func NewArrowRowScanner(resultSetMetadata *cli_service.TGetResultSetMetadataResp
 		return nil, err2
 	}
 
-	var location *time.Location = time.UTC
+	location := time.UTC
 	if cfg != nil {
 		if cfg.Location != nil {
 			location = cfg.Location
diff --git a/internal/rows/arrowbased/arrowRows_test.go b/internal/rows/arrowbased/arrowRows_test.go
index 9d7eaba8..84b27426 100644
--- a/internal/rows/arrowbased/arrowRows_test.go
+++ b/internal/rows/arrowbased/arrowRows_test.go
@@ -239,7 +239,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 		d, _ := NewArrowRowScanner(metadataResp, rowSet, nil, nil, context.Background())
 
-		var ars *arrowRowScanner = d.(*arrowRowScanner)
+		ars := d.(*arrowRowScanner)
 
 		err := ars.makeColumnValuesContainers(ars, rowscanner.NewDelimiter(0, 1))
 		require.Nil(t, err)
@@ -315,7 +315,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 		d, _ := NewArrowRowScanner(metadataResp, rowSet, &cfg, nil, context.Background())
 
-		var ars *arrowRowScanner = d.(*arrowRowScanner)
+		ars := d.(*arrowRowScanner)
 
 		err := ars.makeColumnValuesContainers(ars, rowscanner.NewDelimiter(0, 1))
 		require.Nil(t, err)
@@ -417,7 +417,7 @@ func TestArrowRowScanner(t *testing.T) {
 		d, err1 := NewArrowRowScanner(metadataResp, rowSet, &cfg, nil, context.Background())
 		require.Nil(t, err1)
 
-		var ars *arrowRowScanner = d.(*arrowRowScanner)
+		ars := d.(*arrowRowScanner)
 
 		err := ars.makeColumnValuesContainers(ars, rowscanner.NewDelimiter(0, 0))
 		require.Nil(t, err)
@@ -485,7 +485,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 		d, _ := NewArrowRowScanner(metadataResp, rowSet, &cfg, nil, context.Background())
 
-		var ars *arrowRowScanner = d.(*arrowRowScanner)
+		ars := d.(*arrowRowScanner)
 
 		assert.Nil(t, ars.rowValues)
 
@@ -555,7 +555,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 		d, _ := NewArrowRowScanner(metadataResp, rowSet, &cfg, nil, nil)
 
-		var ars *arrowRowScanner = d.(*arrowRowScanner)
+		ars := d.(*arrowRowScanner)
 
 		fbi := &fakeBatchIterator{
 			batches: []SparkArrowBatch{
@@ -593,7 +593,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 		d, _ := NewArrowRowScanner(metadataResp, rowSet, &cfg, nil, nil)
 
-		var ars *arrowRowScanner = d.(*arrowRowScanner)
+		ars := d.(*arrowRowScanner)
 
 		fbi := &fakeBatchIterator{
 			batches: []SparkArrowBatch{
@@ -632,7 +632,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 		d, _ := NewArrowRowScanner(metadataResp, rowSet, &cfg, nil, nil)
 
-		var ars *arrowRowScanner = d.(*arrowRowScanner)
+		ars := d.(*arrowRowScanner)
 
 		fbi := &fakeBatchIterator{
 			batches: []SparkArrowBatch{
@@ -673,7 +673,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 		d, _ := NewArrowRowScanner(metadataResp, rowSet, &cfg, nil, nil)
 
-		var ars *arrowRowScanner = d.(*arrowRowScanner)
+		ars := d.(*arrowRowScanner)
 
 		fbi := &fakeBatchIterator{
 			batches: []SparkArrowBatch{
@@ -709,7 +709,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 		d, _ := NewArrowRowScanner(metadataResp, rowSet, &cfg, nil, context.Background())
 
-		var ars *arrowRowScanner = d.(*arrowRowScanner)
+		ars := d.(*arrowRowScanner)
 
 		fbi := &fakeBatchIterator{
 			batches: []SparkArrowBatch{
@@ -766,7 +766,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 		var scale int32 = 10
 		var precision int32 = 2
-		var columns []*cli_service.TColumnDesc = []*cli_service.TColumnDesc{
+		columns := []*cli_service.TColumnDesc{
 			{
 				ColumnName: "array_col",
 				TypeDesc: &cli_service.TTypeDesc{
@@ -858,7 +858,7 @@ func TestArrowRowScanner(t *testing.T) {
 
 			d, _ := NewArrowRowScanner(metadataResp, rowSet, &cfg, nil, context.Background())
 
-			var ars *arrowRowScanner = d.(*arrowRowScanner)
+			ars := d.(*arrowRowScanner)
 			ars.UseArrowNativeComplexTypes = true
 			ars.UseArrowNativeDecimal = true
 			ars.UseArrowNativeIntervalTypes = true
@@ -1202,7 +1202,7 @@ func TestArrowRowScanner(t *testing.T) {
 		// verify that the returned values for the complex type
 		// columns are valid json strings
 		var foo []any
-		var s string = dest[10].(string)
+		s := dest[10].(string)
 		err := json.Unmarshal([]byte(s), &foo)
 		assert.Nil(t, err)
 
diff --git a/internal/rows/arrowbased/batchloader.go b/internal/rows/arrowbased/batchloader.go
index e12ea4e6..545aa6e7 100644
--- a/internal/rows/arrowbased/batchloader.go
+++ b/internal/rows/arrowbased/batchloader.go
@@ -35,8 +35,8 @@ func NewCloudIPCStreamIterator(
 	cfg *config.Config,
 ) (IPCStreamIterator, dbsqlerr.DBError) {
 	httpClient := http.DefaultClient
-	if cfg.UserConfig.CloudFetchConfig.HTTPClient != nil {
-		httpClient = cfg.UserConfig.CloudFetchConfig.HTTPClient
+	if cfg.HTTPClient != nil {
+		httpClient = cfg.HTTPClient
 	}
 
 	bi := &cloudIPCStreamIterator{
@@ -160,7 +160,7 @@ func (bi *cloudIPCStreamIterator) Next() (io.Reader, error) {
 			link.RowCount,
 		)
 
-		cancelCtx, cancelFn := context.WithCancel(bi.ctx)
+		cancelCtx, cancelFn := context.WithCancel(bi.ctx) //nolint:gosec // cancelFn stored in task and called on completion
 		task := &cloudFetchDownloadTask{
 			ctx:                cancelCtx,
 			cancel:             cancelFn,
@@ -269,7 +269,7 @@ func (cft *cloudFetchDownloadTask) Run() {
 
 		// Read all data into memory before closing
 		buf, err := io.ReadAll(getReader(data, cft.useLz4Compression))
-		data.Close()
+		data.Close() //nolint:errcheck,gosec // G104: close after reading data
 		if err != nil {
 			cft.resultChan <- cloudFetchDownloadTaskResult{data: nil, err: err}
 			return
diff --git a/internal/rows/arrowbased/batchloader_test.go b/internal/rows/arrowbased/batchloader_test.go
index 99538bbc..5b230610 100644
--- a/internal/rows/arrowbased/batchloader_test.go
+++ b/internal/rows/arrowbased/batchloader_test.go
@@ -271,7 +271,7 @@ func TestCloudFetchIterator(t *testing.T) {
 		cfg := config.WithDefaults()
 		cfg.UseLz4Compression = false
 		cfg.MaxDownloadThreads = 1
-		cfg.UserConfig.CloudFetchConfig.HTTPClient = customHTTPClient
+		cfg.HTTPClient = customHTTPClient
 
 		bi, err := NewCloudBatchIterator(
 			context.Background(),
@@ -309,7 +309,7 @@ func TestCloudFetchIterator(t *testing.T) {
 		cfg.UseLz4Compression = false
 		cfg.MaxDownloadThreads = 1
 		// Explicitly set HTTPClient to nil to verify fallback behavior
-		cfg.UserConfig.CloudFetchConfig.HTTPClient = nil
+		cfg.HTTPClient = nil
 
 		bi, err := NewCloudBatchIterator(
 			context.Background(),
diff --git a/internal/rows/arrowbased/columnValues.go b/internal/rows/arrowbased/columnValues.go
index 0b6fc7d8..47105095 100644
--- a/internal/rows/arrowbased/columnValues.go
+++ b/internal/rows/arrowbased/columnValues.go
@@ -53,7 +53,7 @@ func (rv *rowValues) SetColumnValues(columnIndex int, values arrow.ArrayData) er
 }
 
 func (rv *rowValues) IsNull(columnIndex int, rowNumber int64) bool {
-	var b bool = true
+	b := true
 	if columnIndex < len(rv.columnValueHolders) {
 		b = rv.columnValueHolders[columnIndex].IsNull(int(rowNumber - rv.Start()))
 	}
diff --git a/internal/rows/columnbased/columnRows.go b/internal/rows/columnbased/columnRows.go
index cfaa0daa..b77f230b 100644
--- a/internal/rows/columnbased/columnRows.go
+++ b/internal/rows/columnbased/columnRows.go
@@ -37,7 +37,7 @@ func NewColumnRowScanner(schema *cli_service.TTableSchema, rowSet *cli_service.T
 		logger = dbsqllog.Logger
 	}
 
-	var location *time.Location = time.UTC
+	location := time.UTC
 	if cfg != nil {
 		if cfg.Location != nil {
 			location = cfg.Location
diff --git a/internal/rows/rows.go b/internal/rows/rows.go
index 963a3ce1..7ee2db44 100644
--- a/internal/rows/rows.go
+++ b/internal/rows/rows.go
@@ -57,6 +57,11 @@ type rows struct {
 	logger_ *dbsqllog.DBSQLLogger
 
 	ctx context.Context
+
+	// Telemetry tracking
+	telemetryUpdate func(chunkCount int, bytesDownloaded int64)
+	chunkCount      int
+	bytesDownloaded int64
 }
 
 var _ driver.Rows = (*rows)(nil)
@@ -72,6 +77,7 @@ func NewRows(
 	client cli_service.TCLIService,
 	config *config.Config,
 	directResults *cli_service.TSparkDirectResults,
+	telemetryUpdate func(chunkCount int, bytesDownloaded int64),
 ) (driver.Rows, dbsqlerr.DBError) {
 
 	connId := driverctx.ConnIdFromContext(ctx)
@@ -91,7 +97,7 @@ func NewRows(
 	}
 
 	var pageSize int64 = 10000
-	var location *time.Location = time.UTC
+	location := time.UTC
 	if config != nil {
 		pageSize = int64(config.MaxRows)
 
@@ -103,14 +109,17 @@ func NewRows(
 	logger.Debug().Msgf("databricks: creating Rows, pageSize: %d, location: %v", pageSize, location)
 
 	r := &rows{
-		client:        client,
-		opHandle:      opHandle,
-		connId:        connId,
-		correlationId: correlationId,
-		location:      location,
-		config:        config,
-		logger_:       logger,
-		ctx:           ctx,
+		client:          client,
+		opHandle:        opHandle,
+		connId:          connId,
+		correlationId:   correlationId,
+		location:        location,
+		config:          config,
+		logger_:         logger,
+		ctx:             ctx,
+		telemetryUpdate: telemetryUpdate,
+		chunkCount:      0,
+		bytesDownloaded: 0,
 	}
 
 	// if we already have results for the query do some additional initialization
@@ -127,6 +136,17 @@ func NewRows(
 		if err != nil {
 			return r, err
 		}
+
+		r.chunkCount++
+		if directResults.ResultSet != nil && directResults.ResultSet.Results != nil && directResults.ResultSet.Results.ArrowBatches != nil {
+			for _, batch := range directResults.ResultSet.Results.ArrowBatches {
+				r.bytesDownloaded += int64(len(batch.Batch))
+			}
+		}
+
+		if r.telemetryUpdate != nil {
+			r.telemetryUpdate(r.chunkCount, r.bytesDownloaded)
+		}
 	}
 
 	var d rowscanner.Delimiter
@@ -433,7 +453,7 @@ func (r *rows) getResultSetSchema() (*cli_service.TTableSchema, dbsqlerr.DBError
 
 // fetchResultPage will fetch the result page containing the next row, if necessary
 func (r *rows) fetchResultPage() error {
-	var err dbsqlerr.DBError = isValidRows(r)
+	err := isValidRows(r)
 	if err != nil {
 		return err
 	}
@@ -458,6 +478,19 @@ func (r *rows) fetchResultPage() error {
 		return err1
 	}
 
+	r.chunkCount++
+	if fetchResult != nil && fetchResult.Results != nil {
+		if fetchResult.Results.ArrowBatches != nil {
+			for _, batch := range fetchResult.Results.ArrowBatches {
+				r.bytesDownloaded += int64(len(batch.Batch))
+			}
+		}
+	}
+
+	if r.telemetryUpdate != nil {
+		r.telemetryUpdate(r.chunkCount, r.bytesDownloaded)
+	}
+
 	err1 = r.makeRowScanner(fetchResult)
 	if err1 != nil {
 		return err1
diff --git a/internal/rows/rows_test.go b/internal/rows/rows_test.go
index fa6c913a..bb8ac196 100644
--- a/internal/rows/rows_test.go
+++ b/internal/rows/rows_test.go
@@ -421,11 +421,11 @@ func TestColumnsWithDirectResults(t *testing.T) {
 	ctx := driverctx.NewContextWithConnId(context.Background(), "connId")
 	ctx = driverctx.NewContextWithCorrelationId(ctx, "corrId")
 
-	d, err := NewRows(ctx, nil, client, nil, nil)
+	d, err := NewRows(ctx, nil, client, nil, nil, nil)
 	assert.Nil(t, err)
 
 	rowSet := d.(*rows)
-	defer rowSet.Close()
+	defer rowSet.Close() //nolint:errcheck
 
 	req2 := &cli_service.TGetResultSetMetadataReq{}
 	metadata, _ := client.GetResultSetMetadata(context.Background(), req2)
@@ -720,7 +720,7 @@ func TestRowsCloseOptimization(t *testing.T) {
 	ctx := driverctx.NewContextWithConnId(context.Background(), "connId")
 	ctx = driverctx.NewContextWithCorrelationId(ctx, "corrId")
 	opHandle := &cli_service.TOperationHandle{OperationId: &cli_service.THandleIdentifier{GUID: []byte{'f', 'o'}}}
-	rowSet, _ := NewRows(ctx, opHandle, client, nil, nil)
+	rowSet, _ := NewRows(ctx, opHandle, client, nil, nil, nil)
 
 	// rowSet has no direct results calling Close should result in call to client to close operation
 	err := rowSet.Close()
@@ -733,7 +733,7 @@ func TestRowsCloseOptimization(t *testing.T) {
 		ResultSet:         &cli_service.TFetchResultsResp{Results: &cli_service.TRowSet{Columns: []*cli_service.TColumn{}}},
 	}
 	closeCount = 0
-	rowSet, _ = NewRows(ctx, opHandle, client, nil, directResults)
+	rowSet, _ = NewRows(ctx, opHandle, client, nil, directResults, nil)
 	err = rowSet.Close()
 	assert.Nil(t, err, "rows.Close should not throw an error")
 	assert.Equal(t, 1, closeCount)
@@ -746,7 +746,7 @@ func TestRowsCloseOptimization(t *testing.T) {
 		ResultSetMetadata: &cli_service.TGetResultSetMetadataResp{Schema: &cli_service.TTableSchema{}},
 		ResultSet:         &cli_service.TFetchResultsResp{Results: &cli_service.TRowSet{Columns: []*cli_service.TColumn{}}},
 	}
-	rowSet, _ = NewRows(ctx, opHandle, client, nil, directResults)
+	rowSet, _ = NewRows(ctx, opHandle, client, nil, directResults, nil)
 	err = rowSet.Close()
 	assert.Nil(t, err, "rows.Close should not throw an error")
 	assert.Equal(t, 0, closeCount)
@@ -816,7 +816,7 @@ func TestGetArrowBatches(t *testing.T) {
 
 		client := getSimpleClient([]cli_service.TFetchResultsResp{fetchResp1, fetchResp2})
 		cfg := config.WithDefaults()
-		rows, err := NewRows(ctx, nil, client, cfg, executeStatementResp.DirectResults)
+		rows, err := NewRows(ctx, nil, client, cfg, executeStatementResp.DirectResults, nil)
 		assert.Nil(t, err)
 
 		rows2, ok := rows.(dbsqlrows.Rows)
@@ -889,7 +889,7 @@ func TestGetArrowBatches(t *testing.T) {
 
 		client := getSimpleClient([]cli_service.TFetchResultsResp{fetchResp1, fetchResp2, fetchResp3})
 		cfg := config.WithDefaults()
-		rows, err := NewRows(ctx, nil, client, cfg, nil)
+		rows, err := NewRows(ctx, nil, client, cfg, nil, nil)
 		assert.Nil(t, err)
 
 		rows2, ok := rows.(dbsqlrows.Rows)
@@ -950,7 +950,7 @@ func TestGetArrowBatches(t *testing.T) {
 
 		client := getSimpleClient([]cli_service.TFetchResultsResp{fetchResp1})
 		cfg := config.WithDefaults()
-		rows, err := NewRows(ctx, nil, client, cfg, nil)
+		rows, err := NewRows(ctx, nil, client, cfg, nil, nil)
 		assert.Nil(t, err)
 
 		rows2, ok := rows.(dbsqlrows.Rows)
@@ -977,7 +977,7 @@ func TestGetArrowBatches(t *testing.T) {
 
 		client := getSimpleClient([]cli_service.TFetchResultsResp{})
 		cfg := config.WithDefaults()
-		rows, err := NewRows(ctx, nil, client, cfg, executeStatementResp.DirectResults)
+		rows, err := NewRows(ctx, nil, client, cfg, executeStatementResp.DirectResults, nil)
 		assert.Nil(t, err)
 
 		rows2, ok := rows.(dbsqlrows.Rows)
@@ -1381,17 +1381,18 @@ func getRowsTestSimpleClient(getMetadataCount, fetchResultsCount *int) cli_servi
 
 	fetchResults := func(ctx context.Context, req *cli_service.TFetchResultsReq) (_r *cli_service.TFetchResultsResp, _err error) {
 		*fetchResultsCount++
-		if req.Orientation == cli_service.TFetchOrientation_FETCH_NEXT {
+		switch req.Orientation {
+		case cli_service.TFetchOrientation_FETCH_NEXT:
 			if pageIndex+1 >= len(pages) {
 				return nil, errors.New("can't fetch past end of result set")
 			}
 			pageIndex++
-		} else if req.Orientation == cli_service.TFetchOrientation_FETCH_PRIOR {
+		case cli_service.TFetchOrientation_FETCH_PRIOR:
 			if pageIndex-1 < 0 {
 				return nil, errors.New("can't fetch prior to start of result set")
 			}
 			pageIndex--
-		} else {
+		default:
 			return nil, errors.New("invalid fetch results orientation")
 		}
 
@@ -1556,7 +1557,7 @@ func TestFetchResultPage_PropagatesGetNextPageError(t *testing.T) {
 
 	executeStatementResp := cli_service.TExecuteStatementResp{}
 	cfg := config.WithDefaults()
-	rows, _ := NewRows(ctx, nil, client, cfg, executeStatementResp.DirectResults)
+	rows, _ := NewRows(ctx, nil, client, cfg, executeStatementResp.DirectResults, nil)
 	// Call Next and ensure it propagates the error from getNextPage
 	actualErr := rows.Next(nil)
 
diff --git a/internal/rows/rowscanner/resultPageIterator.go b/internal/rows/rowscanner/resultPageIterator.go
index 144d7447..9cde0950 100644
--- a/internal/rows/rowscanner/resultPageIterator.go
+++ b/internal/rows/rowscanner/resultPageIterator.go
@@ -125,7 +125,7 @@ func (rpf *resultPageIterator) HasNext() bool {
 	if rpf.nextResultPage == nil {
 		nrp, err := rpf.getNextPage()
 		if err != nil {
-			rpf.Close()
+			rpf.Close() //nolint:errcheck,gosec // G104: close in error path
 			rpf.isFinished = true
 			rpf.err = err
 			return false
@@ -134,7 +134,7 @@ func (rpf *resultPageIterator) HasNext() bool {
 		rpf.err = nil
 		rpf.nextResultPage = nrp
 		if !nrp.GetHasMoreRows() {
-			rpf.Close()
+			rpf.Close() //nolint:errcheck,gosec // G104: close in error path
 		}
 	}
 
@@ -285,17 +285,18 @@ func CountRows(rowSet *cli_service.TRowSet) int64 {
 
 // Check if trying to fetch in the specified direction creates an error condition.
 func (rpf *resultPageIterator) checkDirectionValid(direction Direction) error {
-	if direction == DirBack {
+	switch direction {
+	case DirBack:
 		// can't fetch rows previous to the start
 		if rpf.Start() == 0 {
 			return dbsqlerrint.NewDriverError(rpf.ctx, ErrRowsFetchPriorToStart, nil)
 		}
-	} else if direction == DirForward {
+	case DirForward:
 		// can't fetch past the end of the query results
 		if rpf.isFinished {
 			return io.EOF
 		}
-	} else {
+	default:
 		rpf.logger.Error().Msgf(errRowsUnandledFetchDirection(direction.String()))
 		return dbsqlerrint.NewDriverError(rpf.ctx, errRowsUnandledFetchDirection(direction.String()), nil)
 	}
diff --git a/internal/rows/rowscanner/rowScanner.go b/internal/rows/rowscanner/rowScanner.go
index c83cc952..ebb25bf6 100644
--- a/internal/rows/rowscanner/rowScanner.go
+++ b/internal/rows/rowscanner/rowScanner.go
@@ -46,7 +46,7 @@ func IsNull(nulls []byte, position int64) bool {
 	index := position / 8
 	if int64(len(nulls)) > index {
 		b := nulls[index]
-		return (b & (1 << (uint)(position%8))) != 0
+		return (b & (1 << (uint)(position%8))) != 0 //nolint:gosec // position%8 is always 0-7
 	}
 	return false
 }
diff --git a/telemetry/DESIGN.md b/telemetry/DESIGN.md
index 3e742b67..5996ed2e 100644
--- a/telemetry/DESIGN.md
+++ b/telemetry/DESIGN.md
@@ -1,2257 +1,106 @@
-# Databricks SQL Go Driver: Telemetry Design
+# Telemetry Design
 
-## Executive Summary
-
-This document outlines a **telemetry design** for the Databricks SQL Go driver that collects usage metrics and exports them to the Databricks telemetry service. The design leverages Go's `context.Context` and middleware patterns to instrument driver operations without impacting performance.
-
-**Important Note:** Telemetry is **disabled by default** and requires explicit opt-in. A gradual rollout strategy will be used to ensure reliability and user control.
-
-**Key Objectives:**
-- Collect driver usage metrics and performance data
-- Export aggregated metrics to Databricks telemetry service
-- Maintain server-side feature flag control
-- Ensure zero impact on driver performance and reliability
-- Follow Go best practices and idiomatic patterns
-
-**Design Principles:**
-- **Opt-in first**: User explicit consent required, disabled by default
-- **Non-blocking**: All operations async using goroutines
-- **Privacy-first**: No PII or query data collected
-- **Server-controlled**: Feature flag support for gradual rollout
-- **Fail-safe**: All telemetry errors swallowed silently
-- **Idiomatic Go**: Use standard library patterns and interfaces
-
-**Production Requirements** (from JDBC driver experience):
-- **Feature flag caching**: Per-host caching to avoid rate limiting
-- **Circuit breaker**: Protect against telemetry endpoint failures
-- **Exception swallowing**: All telemetry errors caught with minimal logging
-- **Per-host telemetry client**: One client per host to prevent rate limiting
-- **Graceful shutdown**: Proper cleanup with reference counting
-- **Smart exception flushing**: Only flush terminal errors immediately
-
----
-
-## Table of Contents
-
-1. [Background & Motivation](#1-background--motivation)
-2. [Architecture Overview](#2-architecture-overview)
-3. [Core Components](#3-core-components)
-   - 3.1 [featureFlagCache (Per-Host)](#31-featureflagcache-per-host)
-   - 3.2 [clientManager (Per-Host)](#32-clientmanager-per-host)
-   - 3.3 [circuitBreaker](#33-circuitbreaker)
-   - 3.4 [Telemetry Interceptor](#34-telemetry-interceptor)
-   - 3.5 [metricsAggregator](#35-metricsaggregator)
-   - 3.6 [telemetryExporter](#36-telemetryexporter)
-4. [Data Collection](#4-data-collection)
-5. [Export Mechanism](#5-export-mechanism)
-6. [Configuration](#6-configuration)
-   - 6.1 [Configuration Structure](#61-configuration-structure)
-   - 6.2 [Configuration from DSN](#62-configuration-from-dsn)
-   - 6.3 [Feature Flag Integration](#63-feature-flag-integration)
-   - 6.4 [Opt-In Control & Priority](#64-opt-in-control--priority)
-7. [Privacy & Compliance](#7-privacy--compliance)
-8. [Error Handling](#8-error-handling)
-9. [Graceful Shutdown](#9-graceful-shutdown)
-10. [Testing Strategy](#10-testing-strategy)
-11. [Partial Launch Strategy](#11-partial-launch-strategy)
-12. [Implementation Checklist](#12-implementation-checklist)
-13. [References](#13-references)
-
----
-
-## 1. Background & Motivation
-
-### 1.1 Current State
-
-The Databricks SQL Go driver (`databricks-sql-go`) provides:
-- ✅ Standard `database/sql/driver` interface implementation
-- ✅ Thrift-based communication with Databricks clusters
-- ✅ Connection pooling via `database/sql` package
-- ✅ Context-based cancellation and timeouts
-- ✅ Structured logging
-
-### 1.2 Requirements
-
-Product teams need telemetry data to:
-- Understand driver feature adoption (CloudFetch, LZ4 compression, etc.)
-- Identify performance bottlenecks
-- Track error rates and failure modes
-- Make data-driven decisions for driver improvements
-
-### 1.3 Design Goals
-
-**Instrumentation Strategy**:
-- ✅ Minimal code changes to existing driver
-- ✅ Use Go idiomatic patterns (interfaces, middleware, context)
-- ✅ Zero performance impact when telemetry disabled
-- ✅ < 1% overhead when telemetry enabled
-- ✅ No risk to driver reliability
-
----
-
-## 2. Architecture Overview
-
-### 2.1 High-Level Architecture
-
-```mermaid
-graph TB
-    A[Driver Operations] -->|context.Context| B[Telemetry Interceptor]
-    B -->|Collect Metrics| C[metricsAggregator]
-    C -->|Batch & Buffer| D[clientManager]
-    D -->|Get Per-Host Client| E[telemetryClient per Host]
-    E -->|Check Circuit Breaker| F[circuitBreaker]
-    F -->|HTTP POST| G[telemetryExporter]
-    G --> H[Databricks Service]
-    H --> I[Lumberjack]
-
-    J[featureFlagCache per Host] -.->|Enable/Disable| B
-    K[Connection Open] -->|Increment RefCount| D
-    K -->|Increment RefCount| J
-    L[Connection Close] -->|Decrement RefCount| D
-    L -->|Decrement RefCount| J
-
-    style B fill:#e1f5fe
-    style C fill:#e1f5fe
-    style D fill:#ffe0b2
-    style E fill:#ffe0b2
-    style F fill:#ffccbc
-    style J fill:#c8e6c9
-```
-
-**Key Components:**
-1. **Telemetry Interceptor** (new): Wraps driver operations to collect metrics
-2. **featureFlagCache** (new): Per-host caching of feature flags with reference counting
-3. **clientManager** (new): Manages one telemetry client per host with reference counting
-4. **circuitBreaker** (new): Protects against failing telemetry endpoint
-5. **metricsAggregator** (new): Aggregates by statement, batches events
-6. **telemetryExporter** (new): Exports to Databricks service
-
-### 2.2 Data Flow
-
-```mermaid
-sequenceDiagram
-    participant App as Application
-    participant Driver as dbsql.conn
-    participant Interceptor as Telemetry Interceptor
-    participant Agg as metricsAggregator
-    participant Exp as telemetryExporter
-    participant Service as Databricks Service
-
-    App->>Driver: QueryContext(ctx, query)
-    Driver->>Interceptor: Before Execute
-    Interceptor->>Interceptor: Start timer
-
-    Driver->>Driver: Execute operation
-    Driver->>Interceptor: After Execute (with result/error)
-
-    Interceptor->>Agg: RecordMetric(ctx, metric)
-    Agg->>Agg: Aggregate by statement_id
-
-    alt Batch threshold reached
-        Agg->>Exp: Flush(batch)
-        Exp->>Service: POST /telemetry-ext
-    end
-```
-
----
-
-## 3. Core Components
-
-### 3.1 featureFlagCache (Per-Host)
-
-**Purpose**: Cache feature flag values at the host level to avoid repeated API calls and rate limiting.
-
-**Location**: `telemetry/featureflag.go`
-
-#### Rationale
-- **Per-host caching**: Feature flags cached by host to prevent rate limiting
-- **Multi-flag support**: Fetches all flags in a single request for efficiency
-- **Reference counting**: Tracks number of connections per host for proper cleanup
-- **Automatic expiration**: Refreshes cached flags after TTL expires (15 minutes)
-- **Thread-safe**: Uses sync.RWMutex for concurrent access
-- **Synchronous fetch**: Blocks on cache miss (see Section 6.3 for behavior details)
-- **Thundering herd protection**: Only one fetch per host at a time
-
-#### Interface
-
-```go
-package telemetry
-
-import (
-	"context"
-	"net/http"
-	"sync"
-	"time"
-)
-
-// featureFlagCache manages feature flag state per host with reference counting.
-// This prevents rate limiting by caching feature flag responses.
-type featureFlagCache struct {
-	mu       sync.RWMutex
-	contexts map[string]*featureFlagContext
-}
-
-// featureFlagContext holds feature flag state and reference count for a host.
-type featureFlagContext struct {
-	enabled      *bool
-	lastFetched  time.Time
-	refCount     int
-	cacheDuration time.Duration
-}
-
-var (
-	flagCacheOnce     sync.Once
-	flagCacheInstance *featureFlagCache
-)
-
-// getFeatureFlagCache returns the singleton instance.
-func getFeatureFlagCache() *featureFlagCache {
-	flagCacheOnce.Do(func() {
-		flagCacheInstance = &featureFlagCache{
-			contexts: make(map[string]*featureFlagContext),
-		}
-	})
-	return flagCacheInstance
-}
-
-// getOrCreateContext gets or creates a feature flag context for the host.
-// Increments reference count.
-func (c *featureFlagCache) getOrCreateContext(host string) *featureFlagContext {
-	c.mu.Lock()
-	defer c.mu.Unlock()
-
-	ctx, exists := c.contexts[host]
-	if !exists {
-		ctx = &featureFlagContext{
-			cacheDuration: 15 * time.Minute,
-		}
-		c.contexts[host] = ctx
-	}
-	ctx.refCount++
-	return ctx
-}
-
-// releaseContext decrements reference count for the host.
-// Removes context when ref count reaches zero.
-func (c *featureFlagCache) releaseContext(host string) {
-	c.mu.Lock()
-	defer c.mu.Unlock()
-
-	if ctx, exists := c.contexts[host]; exists {
-		ctx.refCount--
-		if ctx.refCount <= 0 {
-			delete(c.contexts, host)
-		}
-	}
-}
-
-// isTelemetryEnabled checks if telemetry is enabled for the host.
-// Uses cached value if available and not expired.
-func (c *featureFlagCache) isTelemetryEnabled(ctx context.Context, host string, httpClient *http.Client) (bool, error) {
-	c.mu.RLock()
-	flagCtx, exists := c.contexts[host]
-	c.mu.RUnlock()
-
-	if !exists {
-		return false, nil
-	}
-
-	// Check if cache is valid
-	if flagCtx.enabled != nil && time.Since(flagCtx.lastFetched) < flagCtx.cacheDuration {
-		return *flagCtx.enabled, nil
-	}
-
-	// Fetch fresh value
-	enabled, err := fetchFeatureFlag(ctx, host, httpClient)
-	if err != nil {
-		// Return cached value on error, or false if no cache
-		if flagCtx.enabled != nil {
-			return *flagCtx.enabled, nil
-		}
-		return false, err
-	}
-
-	// Update cache
-	c.mu.Lock()
-	flagCtx.enabled = &enabled
-	flagCtx.lastFetched = time.Now()
-	c.mu.Unlock()
-
-	return enabled, nil
-}
-
-// isExpired returns true if the cache has expired.
-func (c *featureFlagContext) isExpired() bool {
-	return c.enabled == nil || time.Since(c.lastFetched) > c.cacheDuration
-}
-```
-
----
-
-### 3.2 clientManager (Per-Host)
-
-**Purpose**: Manage one telemetry client per host to prevent rate limiting from concurrent connections.
-
-**Location**: `telemetry/manager.go`
-
-#### Rationale
-- **One client per host**: Large customers open many parallel connections to the same host
-- **Prevents rate limiting**: Shared client batches events from all connections
-- **Reference counting**: Tracks active connections, only closes client when last connection closes
-- **Thread-safe**: Safe for concurrent access from multiple goroutines
-
-#### Interface
-
-```go
-package telemetry
-
-import (
-	"net/http"
-	"sync"
-)
-
-// clientManager manages one telemetry client per host.
-// Prevents rate limiting by sharing clients across connections.
-type clientManager struct {
-	mu      sync.RWMutex
-	clients map[string]*clientHolder
-}
-
-// clientHolder holds a telemetry client and its reference count.
-type clientHolder struct {
-	client   *telemetryClient
-	refCount int
-}
-
-var (
-	managerOnce     sync.Once
-	managerInstance *clientManager
-)
-
-// getClientManager returns the singleton instance.
-func getClientManager() *clientManager {
-	managerOnce.Do(func() {
-		managerInstance = &clientManager{
-			clients: make(map[string]*clientHolder),
-		}
-	})
-	return managerInstance
-}
-
-// getOrCreateClient gets or creates a telemetry client for the host.
-// Increments reference count.
-func (m *clientManager) getOrCreateClient(host string, httpClient *http.Client, cfg *config) *telemetryClient {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	holder, exists := m.clients[host]
-	if !exists {
-		holder = &clientHolder{
-			client: newTelemetryClient(host, httpClient, cfg),
-		}
-		m.clients[host] = holder
-		holder.client.start() // Start background flush goroutine
-	}
-	holder.refCount++
-	return holder.client
-}
-
-// releaseClient decrements reference count for the host.
-// Closes and removes client when ref count reaches zero.
-func (m *clientManager) releaseClient(host string) error {
-	m.mu.Lock()
-	holder, exists := m.clients[host]
-	if !exists {
-		m.mu.Unlock()
-		return nil
-	}
-
-	holder.refCount--
-	if holder.refCount <= 0 {
-		delete(m.clients, host)
-		m.mu.Unlock()
-		return holder.client.close() // Close and flush
-	}
-
-	m.mu.Unlock()
-	return nil
-}
-```
-
----
-
-### 3.3 circuitBreaker
-
-**Purpose**: Implement circuit breaker pattern to protect against failing telemetry endpoint using a sliding window and failure rate percentage algorithm (matching JDBC's Resilience4j implementation).
-
-**Location**: `telemetry/circuitbreaker.go`
-
-#### Rationale
-- **Endpoint protection**: The telemetry endpoint itself may fail or become unavailable
-- **Not just rate limiting**: Protects against 5xx errors, timeouts, network failures
-- **Resource efficiency**: Prevents wasting resources on a failing endpoint
-- **Auto-recovery**: Automatically detects when endpoint becomes healthy again
-- **JDBC alignment**: Uses sliding window with failure rate percentage, matching JDBC driver behavior exactly
-
-#### Algorithm: Sliding Window with Failure Rate
-The circuit breaker tracks recent calls in a **sliding window** (ring buffer) and calculates the **failure rate percentage**:
-- Tracks the last N calls (default: 30)
-- Opens circuit when failure rate >= threshold (default: 50%)
-- Requires minimum calls before evaluation (default: 20)
-- Uses percentage-based evaluation instead of consecutive failures
-
-**Example**: With 30 calls in window, if 15 or more fail (50%), circuit opens. This is more robust than consecutive-failure counting as it considers overall reliability.
-
-#### States
-1. **Closed**: Normal operation, requests pass through
-2. **Open**: After failure rate exceeds threshold, all requests rejected immediately (drop events)
-3. **Half-Open**: After wait duration, allows test requests to check if endpoint recovered
-
-#### Configuration (matching JDBC defaults)
-- **failureRateThreshold**: 50% - Opens circuit if failure rate >= 50%
-- **minimumNumberOfCalls**: 20 - Minimum calls before evaluating failure rate
-- **slidingWindowSize**: 30 - Track last 30 calls in sliding window
-- **waitDurationInOpenState**: 30s - Wait before transitioning to half-open
-- **permittedCallsInHalfOpen**: 3 - Test with 3 successful calls before closing
-
-#### Interface
-
-```go
-package telemetry
-
-import (
-	"context"
-	"errors"
-	"sync"
-	"sync/atomic"
-	"time"
-)
-
-// circuitState represents the state of the circuit breaker.
-type circuitState int32
-
-const (
-	stateClosed circuitState = iota
-	stateOpen
-	stateHalfOpen
-)
-
-// callResult represents the result of a call (success or failure).
-type callResult bool
-
-const (
-	callSuccess callResult = true
-	callFailure callResult = false
-)
-
-// circuitBreaker implements the circuit breaker pattern.
-// It protects against failing telemetry endpoints by tracking failures
-// using a sliding window and failure rate percentage.
-type circuitBreaker struct {
-	mu sync.RWMutex
-
-	state         atomic.Int32 // circuitState
-	lastStateTime time.Time
-
-	// Sliding window for tracking calls
-	window       []callResult
-	windowIndex  int
-	windowFilled bool
-	totalCalls   int
-	failureCount int
-
-	// Half-open state tracking
-	halfOpenSuccesses int
-
-	config circuitBreakerConfig
-}
-
-// circuitBreakerConfig holds circuit breaker configuration.
-type circuitBreakerConfig struct {
-	failureRateThreshold     int           // Open if failure rate >= this percentage (0-100)
-	minimumNumberOfCalls     int           // Minimum calls before evaluating failure rate
-	slidingWindowSize        int           // Number of recent calls to track
-	waitDurationInOpenState  time.Duration // Wait before transitioning to half-open
-	permittedCallsInHalfOpen int           // Number of test calls in half-open state
-}
-
-// defaultCircuitBreakerConfig returns default configuration matching JDBC.
-func defaultCircuitBreakerConfig() circuitBreakerConfig {
-	return circuitBreakerConfig{
-		failureRateThreshold:     50, // 50% failure rate
-		minimumNumberOfCalls:     20, // Minimum sample size
-		slidingWindowSize:        30, // Keep recent 30 calls
-		waitDurationInOpenState:  30 * time.Second,
-		permittedCallsInHalfOpen: 3, // Test with 3 calls
-	}
-}
-
-// newCircuitBreaker creates a new circuit breaker.
-func newCircuitBreaker(cfg circuitBreakerConfig) *circuitBreaker {
-	cb := &circuitBreaker{
-		config:        cfg,
-		lastStateTime: time.Now(),
-		window:        make([]callResult, cfg.slidingWindowSize),
-	}
-	cb.state.Store(int32(stateClosed))
-	return cb
-}
-
-// ErrCircuitOpen is returned when circuit is open.
-var ErrCircuitOpen = errors.New("circuit breaker is open")
-
-// execute executes the function if circuit allows.
-func (cb *circuitBreaker) execute(ctx context.Context, fn func() error) error {
-	state := circuitState(cb.state.Load())
-
-	switch state {
-	case stateOpen:
-		// Check if wait duration has passed
-		cb.mu.RLock()
-		shouldRetry := time.Since(cb.lastStateTime) > cb.config.waitDurationInOpenState
-		cb.mu.RUnlock()
-
-		if shouldRetry {
-			// Transition to half-open
-			cb.setState(stateHalfOpen)
-			return cb.tryExecute(ctx, fn)
-		}
-		return ErrCircuitOpen
-
-	case stateHalfOpen:
-		return cb.tryExecute(ctx, fn)
-
-	case stateClosed:
-		return cb.tryExecute(ctx, fn)
-	}
-
-	return nil
-}
-
-// tryExecute attempts to execute the function and updates state.
-func (cb *circuitBreaker) tryExecute(ctx context.Context, fn func() error) error {
-	err := fn()
-
-	if err != nil {
-		cb.recordCall(callFailure)
-		return err
-	}
-
-	cb.recordCall(callSuccess)
-	return nil
-}
-
-// recordCall records a call result in the sliding window and evaluates state transitions.
-func (cb *circuitBreaker) recordCall(result callResult) {
-	cb.mu.Lock()
-	defer cb.mu.Unlock()
-
-	state := circuitState(cb.state.Load())
-
-	// Handle half-open state specially
-	if state == stateHalfOpen {
-		if result == callFailure {
-			// Any failure in half-open immediately reopens circuit
-			cb.resetWindowUnlocked()
-			cb.setStateUnlocked(stateOpen)
-			return
-		}
-
-		cb.halfOpenSuccesses++
-		if cb.halfOpenSuccesses >= cb.config.permittedCallsInHalfOpen {
-			// Enough successes to close circuit
-			cb.resetWindowUnlocked()
-			cb.setStateUnlocked(stateClosed)
-		}
-		return
-	}
-
-	// Record in sliding window
-	// Remove old value from count if window is full
-	if cb.windowFilled && cb.window[cb.windowIndex] == callFailure {
-		cb.failureCount--
-	}
-
-	// Add new value
-	cb.window[cb.windowIndex] = result
-	if result == callFailure {
-		cb.failureCount++
-	}
-
-	// Move to next position
-	cb.windowIndex = (cb.windowIndex + 1) % cb.config.slidingWindowSize
-	if cb.windowIndex == 0 {
-		cb.windowFilled = true
-	}
-
-	cb.totalCalls++
-
-	// Evaluate if we should open the circuit
-	if state == stateClosed {
-		cb.evaluateStateUnlocked()
-	}
-}
-
-// evaluateStateUnlocked checks if the circuit should open based on failure rate.
-// Caller must hold cb.mu lock.
-func (cb *circuitBreaker) evaluateStateUnlocked() {
-	// Need minimum number of calls before evaluating
-	windowSize := cb.totalCalls
-	if cb.windowFilled {
-		windowSize = cb.config.slidingWindowSize
-	}
-
-	if windowSize < cb.config.minimumNumberOfCalls {
-		return
-	}
-
-	// Calculate failure rate
-	failureRate := (cb.failureCount * 100) / windowSize
-
-	if failureRate >= cb.config.failureRateThreshold {
-		cb.setStateUnlocked(stateOpen)
-	}
-}
-
-// resetWindowUnlocked clears the sliding window.
-// Caller must hold cb.mu lock.
-func (cb *circuitBreaker) resetWindowUnlocked() {
-	cb.windowIndex = 0
-	cb.windowFilled = false
-	cb.totalCalls = 0
-	cb.failureCount = 0
-	cb.halfOpenSuccesses = 0
-}
-
-// setState transitions to a new state.
-func (cb *circuitBreaker) setState(newState circuitState) {
-	cb.mu.Lock()
-	defer cb.mu.Unlock()
-	cb.setStateUnlocked(newState)
-}
-
-// setStateUnlocked transitions to a new state without locking.
-// Caller must hold cb.mu lock.
-func (cb *circuitBreaker) setStateUnlocked(newState circuitState) {
-	oldState := circuitState(cb.state.Load())
-	if oldState == newState {
-		return
-	}
-
-	cb.state.Store(int32(newState))
-	cb.lastStateTime = time.Now()
-
-	// Log state transition at DEBUG level
-	// logger.Debug().Msgf("circuit breaker: %v -> %v", oldState, newState)
-}
-
-// circuitBreakerManager manages circuit breakers per host.
-// Each host gets its own circuit breaker to provide isolation.
-type circuitBreakerManager struct {
-	mu       sync.RWMutex
-	breakers map[string]*circuitBreaker
-}
-
-var (
-	breakerManagerOnce     sync.Once
-	breakerManagerInstance *circuitBreakerManager
-)
-
-// getCircuitBreakerManager returns the singleton instance.
-func getCircuitBreakerManager() *circuitBreakerManager {
-	breakerManagerOnce.Do(func() {
-		breakerManagerInstance = &circuitBreakerManager{
-			breakers: make(map[string]*circuitBreaker),
-		}
-	})
-	return breakerManagerInstance
-}
-
-// getCircuitBreaker gets or creates a circuit breaker for the host.
-// Thread-safe for concurrent access.
-func (m *circuitBreakerManager) getCircuitBreaker(host string) *circuitBreaker {
-	m.mu.RLock()
-	cb, exists := m.breakers[host]
-	m.mu.RUnlock()
-
-	if exists {
-		return cb
-	}
-
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	// Double-check after acquiring write lock
-	if cb, exists = m.breakers[host]; exists {
-		return cb
-	}
-
-	cb = newCircuitBreaker(defaultCircuitBreakerConfig())
-	m.breakers[host] = cb
-	return cb
-}
-```
-
----
-
-### 3.4 Telemetry Interceptor
-
-**Purpose**: Intercept driver operations to collect metrics without modifying core driver logic.
-
-**Location**: `telemetry/interceptor.go`
-
-#### Approach
-
-Use Go's context and function wrapping patterns to intercept operations:
-
-```go
-package telemetry
-
-import (
-	"context"
-	"time"
-)
-
-// interceptor wraps driver operations to collect metrics.
-type interceptor struct {
-	aggregator *metricsAggregator
-	enabled    bool
-}
-
-// metricContext holds metric collection state in context.
-type metricContext struct {
-	statementID string
-	startTime   time.Time
-	tags        map[string]interface{}
-}
-
-type contextKey int
-
-const metricContextKey contextKey = 0
-
-// withMetricContext adds metric context to the context.
-func withMetricContext(ctx context.Context, mc *metricContext) context.Context {
-	return context.WithValue(ctx, metricContextKey, mc)
-}
-
-// getMetricContext retrieves metric context from the context.
-func getMetricContext(ctx context.Context) *metricContext {
-	if mc, ok := ctx.Value(metricContextKey).(*metricContext); ok {
-		return mc
-	}
-	return nil
-}
-
-// beforeExecute is called before statement execution.
-func (i *interceptor) beforeExecute(ctx context.Context, statementID string) context.Context {
-	if !i.enabled {
-		return ctx
-	}
-
-	mc := &metricContext{
-		statementID: statementID,
-		startTime:   time.Now(),
-		tags:        make(map[string]interface{}),
-	}
-
-	return withMetricContext(ctx, mc)
-}
-
-// afterExecute is called after statement execution.
-func (i *interceptor) afterExecute(ctx context.Context, err error) {
-	if !i.enabled {
-		return
-	}
-
-	mc := getMetricContext(ctx)
-	if mc == nil {
-		return
-	}
-
-	// Swallow all panics
-	defer func() {
-		if r := recover(); r != nil {
-			// Log at trace level only
-		}
-	}()
-
-	metric := &telemetryMetric{
-		metricType:  "statement",
-		timestamp:   mc.startTime,
-		statementID: mc.statementID,
-		latencyMs:   time.Since(mc.startTime).Milliseconds(),
-		tags:        mc.tags,
-	}
-
-	if err != nil {
-		metric.errorType = classifyError(err)
-	}
-
-	// Non-blocking send
-	i.aggregator.recordMetric(ctx, metric)
-}
-
-// addTag adds a tag to the current metric context.
-func (i *interceptor) addTag(ctx context.Context, key string, value interface{}) {
-	if !i.enabled {
-		return
-	}
-
-	mc := getMetricContext(ctx)
-	if mc != nil {
-		mc.tags[key] = value
-	}
-}
-```
-
-**Integration points in driver**:
-
-```go
-// In statement.go
-func (s *stmt) QueryContext(ctx context.Context, args []driver.NamedValue) (driver.Rows, error) {
-	// Wrap context with telemetry
-	if s.conn.telemetry != nil {
-		ctx = s.conn.telemetry.beforeExecute(ctx, generateStatementID())
-		defer func() {
-			s.conn.telemetry.afterExecute(ctx, err)
-		}()
-	}
-
-	// Existing implementation
-	// ...
-}
-```
-
----
-
-### 3.5 metricsAggregator
-
-**Purpose**: Aggregate metrics by statement and batch for efficient export.
-
-**Location**: `telemetry/aggregator.go`
-
-#### Interface
-
-```go
-package telemetry
-
-import (
-	"context"
-	"sync"
-	"time"
-)
-
-// metricsAggregator aggregates metrics by statement and batches for export.
-type metricsAggregator struct {
-	mu sync.RWMutex
-
-	statements map[string]*statementMetrics
-	batch      []*telemetryMetric
-	exporter   *telemetryExporter
-
-	batchSize     int
-	flushInterval time.Duration
-	flushTimer    *time.Timer
-	stopCh        chan struct{}
-}
-
-// statementMetrics holds aggregated metrics for a statement.
-type statementMetrics struct {
-	statementID  string
-	sessionID    string
-	totalLatency time.Duration
-	chunkCount   int
-	bytesDownloaded int64
-	pollCount    int
-	errors       []error
-	tags         map[string]interface{}
-}
-
-// newMetricsAggregator creates a new metrics aggregator.
-func newMetricsAggregator(exporter *telemetryExporter, cfg *config) *metricsAggregator {
-	agg := &metricsAggregator{
-		statements:    make(map[string]*statementMetrics),
-		batch:         make([]*telemetryMetric, 0, cfg.batchSize),
-		exporter:      exporter,
-		batchSize:     cfg.batchSize,
-		flushInterval: cfg.flushInterval,
-		stopCh:        make(chan struct{}),
-	}
-
-	// Start background flush timer
-	go agg.flushLoop()
-
-	return agg
-}
-
-// recordMetric records a metric for aggregation.
-func (agg *metricsAggregator) recordMetric(ctx context.Context, metric *telemetryMetric) {
-	// Swallow all errors
-	defer func() {
-		if r := recover(); r != nil {
-			// Log at trace level only
-		}
-	}()
-
-	agg.mu.Lock()
-	defer agg.mu.Unlock()
-
-	switch metric.metricType {
-	case "connection":
-		// Emit immediately
-		agg.batch = append(agg.batch, metric)
-		if len(agg.batch) >= agg.batchSize {
-			agg.flushUnlocked(ctx)
-		}
-
-	case "statement":
-		// Aggregate by statement ID
-		stmt, exists := agg.statements[metric.statementID]
-		if !exists {
-			stmt = &statementMetrics{
-				statementID: metric.statementID,
-				tags:        make(map[string]interface{}),
-			}
-			agg.statements[metric.statementID] = stmt
-		}
-
-		// Update aggregated values
-		stmt.totalLatency += time.Duration(metric.latencyMs) * time.Millisecond
-		if chunkCount, ok := metric.tags["chunk_count"].(int); ok {
-			stmt.chunkCount += chunkCount
-		}
-		if bytes, ok := metric.tags["bytes_downloaded"].(int64); ok {
-			stmt.bytesDownloaded += bytes
-		}
-
-		// Merge tags
-		for k, v := range metric.tags {
-			stmt.tags[k] = v
-		}
-
-	case "error":
-		// Check if terminal
-		if isTerminalError(metric.errorType) {
-			// Flush immediately
-			agg.batch = append(agg.batch, metric)
-			agg.flushUnlocked(ctx)
-		} else {
-			// Buffer with statement
-			if stmt, exists := agg.statements[metric.statementID]; exists {
-				stmt.errors = append(stmt.errors, errors.New(metric.errorType))
-			}
-		}
-	}
-}
-
-// completeStatement marks a statement as complete and emits aggregated metric.
-func (agg *metricsAggregator) completeStatement(ctx context.Context, statementID string, failed bool) {
-	defer func() {
-		if r := recover(); r != nil {
-			// Log at trace level only
-		}
-	}()
-
-	agg.mu.Lock()
-	defer agg.mu.Unlock()
-
-	stmt, exists := agg.statements[statementID]
-	if !exists {
-		return
-	}
-	delete(agg.statements, statementID)
-
-	// Create aggregated metric
-	metric := &telemetryMetric{
-		metricType:      "statement",
-		timestamp:       time.Now(),
-		statementID:     stmt.statementID,
-		sessionID:       stmt.sessionID,
-		latencyMs:       stmt.totalLatency.Milliseconds(),
-		tags:            stmt.tags,
-	}
-
-	// Add aggregated counts
-	metric.tags["chunk_count"] = stmt.chunkCount
-	metric.tags["bytes_downloaded"] = stmt.bytesDownloaded
-	metric.tags["poll_count"] = stmt.pollCount
-
-	agg.batch = append(agg.batch, metric)
-
-	// Emit errors if statement failed
-	if failed && len(stmt.errors) > 0 {
-		for _, err := range stmt.errors {
-			errorMetric := &telemetryMetric{
-				metricType:  "error",
-				timestamp:   time.Now(),
-				statementID: statementID,
-				errorType:   err.Error(),
-			}
-			agg.batch = append(agg.batch, errorMetric)
-		}
-	}
-
-	// Flush if batch full
-	if len(agg.batch) >= agg.batchSize {
-		agg.flushUnlocked(ctx)
-	}
-}
-
-// flushLoop runs periodic flush in background.
-func (agg *metricsAggregator) flushLoop() {
-	ticker := time.NewTicker(agg.flushInterval)
-	defer ticker.Stop()
-
-	for {
-		select {
-		case <-ticker.C:
-			agg.flush(context.Background())
-		case <-agg.stopCh:
-			return
-		}
-	}
-}
-
-// flush flushes pending metrics to exporter.
-func (agg *metricsAggregator) flush(ctx context.Context) {
-	agg.mu.Lock()
-	defer agg.mu.Unlock()
-	agg.flushUnlocked(ctx)
-}
-
-// flushUnlocked flushes without locking (caller must hold lock).
-func (agg *metricsAggregator) flushUnlocked(ctx context.Context) {
-	if len(agg.batch) == 0 {
-		return
-	}
-
-	// Copy batch and clear
-	metrics := make([]*telemetryMetric, len(agg.batch))
-	copy(metrics, agg.batch)
-	agg.batch = agg.batch[:0]
-
-	// Export asynchronously
-	go func() {
-		defer func() {
-			if r := recover(); r != nil {
-				// Log at trace level only
-			}
-		}()
-		agg.exporter.export(ctx, metrics)
-	}()
-}
-
-// close stops the aggregator and flushes pending metrics.
-func (agg *metricsAggregator) close(ctx context.Context) error {
-	close(agg.stopCh)
-	agg.flush(ctx)
-	return nil
-}
-```
+Telemetry is **disabled by default**. It collects driver usage metrics (latency, error types, feature flags) and exports them to the Databricks telemetry service. No SQL text, PII, or query results are ever collected.
 
 ---
 
-### 3.6 telemetryExporter
+## Architecture
 
-**Purpose**: Export metrics to Databricks telemetry service.
-
-**Location**: `telemetry/exporter.go`
-
-#### Interface
-
-```go
-package telemetry
-
-import (
-	"bytes"
-	"context"
-	"encoding/json"
-	"fmt"
-	"net/http"
-	"time"
-)
-
-// telemetryExporter exports metrics to Databricks service.
-type telemetryExporter struct {
-	host          string
-	httpClient    *http.Client
-	circuitBreaker *circuitBreaker
-	cfg           *config
-}
-
-// telemetryMetric represents a metric to export.
-type telemetryMetric struct {
-	metricType      string
-	timestamp       time.Time
-	workspaceID     string
-	sessionID       string
-	statementID     string
-	latencyMs       int64
-	errorType       string
-	tags            map[string]interface{}
-}
-
-// newTelemetryExporter creates a new exporter.
-func newTelemetryExporter(host string, httpClient *http.Client, cfg *config) *telemetryExporter {
-	return &telemetryExporter{
-		host:          host,
-		httpClient:    httpClient,
-		circuitBreaker: getCircuitBreakerManager().getCircuitBreaker(host),
-		cfg:           cfg,
-	}
-}
-
-// export exports metrics to Databricks service.
-func (e *telemetryExporter) export(ctx context.Context, metrics []*telemetryMetric) {
-	// Swallow all errors
-	defer func() {
-		if r := recover(); r != nil {
-			// Log at trace level only
-		}
-	}()
-
-	// Check circuit breaker
-	err := e.circuitBreaker.execute(ctx, func() error {
-		return e.doExport(ctx, metrics)
-	})
-
-	if err == ErrCircuitOpen {
-		// Drop metrics silently
-		return
-	}
-
-	if err != nil {
-		// Log at trace level only
-	}
-}
-
-// doExport performs the actual export with retries.
-func (e *telemetryExporter) doExport(ctx context.Context, metrics []*telemetryMetric) error {
-	// Serialize metrics
-	data, err := json.Marshal(metrics)
-	if err != nil {
-		return fmt.Errorf("failed to marshal metrics: %w", err)
-	}
-
-	// Determine endpoint
-	endpoint := fmt.Sprintf("https://%s/api/2.0/telemetry-ext", e.host)
-
-	// Create request
-	req, err := http.NewRequestWithContext(ctx, "POST", endpoint, bytes.NewReader(data))
-	if err != nil {
-		return fmt.Errorf("failed to create request: %w", err)
-	}
-
-	req.Header.Set("Content-Type", "application/json")
-
-	// Retry logic
-	maxRetries := e.cfg.maxRetries
-	for attempt := 0; attempt <= maxRetries; attempt++ {
-		if attempt > 0 {
-			// Exponential backoff
-			backoff := time.Duration(attempt) * e.cfg.retryDelay
-			time.Sleep(backoff)
-		}
-
-		resp, err := e.httpClient.Do(req)
-		if err != nil {
-			if attempt == maxRetries {
-				return fmt.Errorf("failed after %d retries: %w", maxRetries, err)
-			}
-			continue
-		}
-
-		resp.Body.Close()
-
-		// Check status code
-		if resp.StatusCode >= 200 && resp.StatusCode < 300 {
-			return nil // Success
-		}
-
-		// Check if retryable
-		if !isRetryableStatus(resp.StatusCode) {
-			return fmt.Errorf("non-retryable status: %d", resp.StatusCode)
-		}
-
-		if attempt == maxRetries {
-			return fmt.Errorf("failed after %d retries: status %d", maxRetries, resp.StatusCode)
-		}
-	}
-
-	return nil
-}
-
-// isRetryableStatus returns true if HTTP status is retryable.
-func isRetryableStatus(status int) bool {
-	return status == 429 || status == 503 || status >= 500
-}
 ```
-
----
-
-## 4. Data Collection
-
-### 4.1 Metric Tags
-
-All metric tags are defined in a centralized location for maintainability and security.
-
-**Location**: `telemetry/tags.go`
-
-```go
-package telemetry
-
-// Tag names for connection metrics
-const (
-	TagWorkspaceID   = "workspace.id"
-	TagSessionID     = "session.id"
-	TagDriverVersion = "driver.version"
-	TagDriverOS      = "driver.os"
-	TagDriverRuntime = "driver.runtime"
-	TagServerAddress = "server.address" // Not exported to Databricks
-)
-
-// Tag names for statement metrics
-const (
-	TagStatementID          = "statement.id"
-	TagResultFormat         = "result.format"
-	TagResultChunkCount     = "result.chunk_count"
-	TagResultBytesDownloaded = "result.bytes_downloaded"
-	TagCompressionEnabled   = "result.compression_enabled"
-	TagPollCount            = "poll.count"
-	TagPollLatency          = "poll.latency_ms"
-)
-
-// Tag names for error metrics
-const (
-	TagErrorType = "error.type"
-	TagErrorCode = "error.code"
-)
-
-// Feature flag tags
-const (
-	TagFeatureCloudFetch     = "feature.cloudfetch"
-	TagFeatureLZ4            = "feature.lz4"
-	TagFeatureDirectResults  = "feature.direct_results"
-)
-
-// tagExportScope defines where a tag can be exported.
-type tagExportScope int
-
-const (
-	exportNone tagExportScope = 0
-	exportLocal     = 1 << iota
-	exportDatabricks
-	exportAll = exportLocal | exportDatabricks
-)
-
-// tagDefinition defines a metric tag and its export scope.
-type tagDefinition struct {
-	name        string
-	exportScope tagExportScope
-	description string
-	required    bool
-}
-
-// connectionTags returns tags allowed for connection events.
-func connectionTags() []tagDefinition {
-	return []tagDefinition{
-		{TagWorkspaceID, exportDatabricks, "Databricks workspace ID", true},
-		{TagSessionID, exportDatabricks, "Connection session ID", true},
-		{TagDriverVersion, exportAll, "Driver version", false},
-		{TagDriverOS, exportAll, "Operating system", false},
-		{TagDriverRuntime, exportAll, "Go runtime version", false},
-		{TagFeatureCloudFetch, exportDatabricks, "CloudFetch enabled", false},
-		{TagFeatureLZ4, exportDatabricks, "LZ4 compression enabled", false},
-		{TagServerAddress, exportLocal, "Server address (local only)", false},
-	}
-}
-
-// statementTags returns tags allowed for statement events.
-func statementTags() []tagDefinition {
-	return []tagDefinition{
-		{TagStatementID, exportDatabricks, "Statement ID", true},
-		{TagSessionID, exportDatabricks, "Session ID", true},
-		{TagResultFormat, exportDatabricks, "Result format", false},
-		{TagResultChunkCount, exportDatabricks, "Chunk count", false},
-		{TagResultBytesDownloaded, exportDatabricks, "Bytes downloaded", false},
-		{TagCompressionEnabled, exportDatabricks, "Compression enabled", false},
-		{TagPollCount, exportDatabricks, "Poll count", false},
-		{TagPollLatency, exportDatabricks, "Poll latency", false},
-	}
-}
-
-// shouldExportToDatabricks returns true if tag should be exported to Databricks.
-func shouldExportToDatabricks(metricType, tagName string) bool {
-	var tags []tagDefinition
-	switch metricType {
-	case "connection":
-		tags = connectionTags()
-	case "statement":
-		tags = statementTags()
-	default:
-		return false
-	}
-
-	for _, tag := range tags {
-		if tag.name == tagName {
-			return tag.exportScope&exportDatabricks != 0
-		}
-	}
-	return false
-}
+Driver operations
+    → Interceptor          (collect + tag metrics via context)
+    → metricsAggregator    (aggregate by statement, batch by size/time)
+    → telemetryExporter    (HTTP POST with retries)
+    → circuitBreaker       (protect against endpoint failures)
+    → /telemetry-ext
 ```
 
-### 4.2 Metric Collection Points
-
-```go
-// In connection.go - Connection open
-func (c *connector) Connect(ctx context.Context) (driver.Conn, error) {
-	// ... existing code ...
-
-	conn := &conn{
-		id:      generateConnectionID(),
-		cfg:     c.cfg,
-		client:  tclient,
-		session: session,
-	}
-
-	// Initialize telemetry if enabled
-	if c.cfg.telemetryEnabled {
-		conn.telemetry = newTelemetryInterceptor(conn.id, c.cfg)
-
-		// Record connection open
-		conn.telemetry.recordConnection(ctx, map[string]interface{}{
-			TagWorkspaceID:   c.cfg.WorkspaceID,
-			TagSessionID:     conn.id,
-			TagDriverVersion: c.cfg.DriverVersion,
-			TagDriverOS:      runtime.GOOS,
-			TagDriverRuntime: runtime.Version(),
-		})
-	}
-
-	return conn, nil
-}
-
-// In statement.go - Statement execution
-func (s *stmt) QueryContext(ctx context.Context, args []driver.NamedValue) (driver.Rows, error) {
-	statementID := generateStatementID()
-
-	// Start telemetry
-	if s.conn.telemetry != nil {
-		ctx = s.conn.telemetry.beforeExecute(ctx, statementID)
-		defer func() {
-			s.conn.telemetry.afterExecute(ctx, err)
-			s.conn.telemetry.completeStatement(ctx, statementID, err != nil)
-		}()
-	}
-
-	// ... existing code ...
-
-	// Add telemetry tags during execution
-	if s.conn.telemetry != nil {
-		s.conn.telemetry.addTag(ctx, TagResultFormat, "cloudfetch")
-		s.conn.telemetry.addTag(ctx, TagResultChunkCount, chunkCount)
-	}
-
-	return rows, nil
-}
-```
+One `metricsAggregator` + `telemetryExporter` is shared per host (managed by `clientManager` with reference counting). This prevents rate limiting when many connections open to the same host.
 
 ---
 
-## 5. Export Mechanism
+## Components
 
-### 5.1 Data Model
+### `Interceptor` (`interceptor.go`)
+Exported hooks called by the driver package:
 
-```go
-// telemetryPayload is the JSON structure sent to Databricks.
-type telemetryPayload struct {
-	Metrics []*exportedMetric `json:"metrics"`
-}
-
-// exportedMetric is a single metric in the payload.
-type exportedMetric struct {
-	MetricType   string                 `json:"metric_type"`
-	Timestamp    string                 `json:"timestamp"` // RFC3339
-	WorkspaceID  string                 `json:"workspace_id,omitempty"`
-	SessionID    string                 `json:"session_id,omitempty"`
-	StatementID  string                 `json:"statement_id,omitempty"`
-	LatencyMs    int64                  `json:"latency_ms,omitempty"`
-	ErrorType    string                 `json:"error_type,omitempty"`
-	Tags         map[string]interface{} `json:"tags,omitempty"`
-}
-
-// toExportedMetric converts internal metric to exported format.
-func (m *telemetryMetric) toExportedMetric() *exportedMetric {
-	// Filter tags based on export scope
-	filteredTags := make(map[string]interface{})
-	for k, v := range m.tags {
-		if shouldExportToDatabricks(m.metricType, k) {
-			filteredTags[k] = v
-		}
-	}
+| Method | When called |
+|--------|-------------|
+| `BeforeExecute(ctx, sessionID, statementID)` | Before statement execution |
+| `BeforeExecuteWithTime(ctx, sessionID, statementID, startTime)` | When statement ID is known only after execution starts |
+| `AfterExecute(ctx, err)` | After execution completes |
+| `AddTag(ctx, key, value)` | During execution to attach metadata |
+| `CompleteStatement(ctx, statementID, failed)` | After rows are fully consumed |
+| `RecordOperation(ctx, sessionID, opType, latencyMs)` | For non-statement operations |
+| `Close(ctx)` | On connection close — **synchronous blocking** flush; waits for the HTTP export to complete before returning, matching JDBC's `flush(true).get()` behavior |
 
-	return &exportedMetric{
-		MetricType:  m.metricType,
-		Timestamp:   m.timestamp.Format(time.RFC3339),
-		WorkspaceID: m.workspaceID,
-		SessionID:   m.sessionID,
-		StatementID: m.statementID,
-		LatencyMs:   m.latencyMs,
-		ErrorType:   m.errorType,
-		Tags:        filteredTags,
-	}
-}
-```
+### `metricsAggregator` (`aggregator.go`)
+- Aggregates statement metrics by `statementID` (total latency, chunk count, poll count)
+- Flush behavior by metric type:
+  - `"connection"` — flushes immediately (lifecycle event, must not be lost)
+  - `"operation"` — batched, flushes when batch reaches `BatchSize`
+  - `"statement"` — accumulated until `CompleteStatement`, then batched
+- Background flush ticker runs every `FlushInterval`
+- `flushUnlocked` exports asynchronously; semaphore (8 slots) bounds concurrent goroutines
+- `flushSync` exports synchronously; used by `Interceptor.Close` to block until delivery
+- `close` is idempotent via `sync.Once`; cancels in-flight periodic exports
 
-### 5.2 Export Endpoints
+### `telemetryExporter` (`exporter.go`)
+- HTTP POST to `<host>/telemetry-ext`
+- Exponential backoff retries (default: 3 retries, 100ms base delay)
+- Retryable statuses: 429, 503, 5xx
+- All errors and panics are swallowed; logged at TRACE level only
 
-```go
-const (
-	// telemetryEndpointAuth is used when connection has auth token
-	telemetryEndpointAuth = "/api/2.0/telemetry-ext"
+### `circuitBreaker` (`circuitbreaker.go`)
+Sliding window (30 calls), opens when failure rate ≥ 50% over ≥ 20 calls. Recovers after 30s via half-open state (3 successful test calls to close). One circuit breaker per host.
 
-	// telemetryEndpointUnauth is used when no auth available
-	telemetryEndpointUnauth = "/api/2.0/telemetry-unauth"
-)
+### `clientManager` (`manager.go`)
+Singleton. Maintains one `(aggregator + exporter)` per host with reference counting. Last connection to close triggers aggregator shutdown and final flush.
 
-// selectEndpoint chooses the appropriate endpoint based on auth.
-func (e *telemetryExporter) selectEndpoint() string {
-	if e.cfg.hasAuth {
-		return telemetryEndpointAuth
-	}
-	return telemetryEndpointUnauth
-}
-```
+### `featureFlagCache` (`featureflag.go`)
+Caches the server-side feature flag per host for 15 minutes. Synchronous fetch on first connection (or cache expiry); concurrent callers receive the stale cached value immediately.
 
 ---
 
-## 6. Configuration
-
-### 6.1 Configuration Structure
-
-```go
-package telemetry
-
-import "time"
-
-// Config holds telemetry configuration.
-type Config struct {
-	// Enabled controls whether telemetry is active
-	Enabled bool
-
-	// EnableTelemetry indicates user wants telemetry enabled.
-	// Follows client > server > default priority: if set by the client it takes
-	// precedence; otherwise the server feature flag and defaults are consulted.
-	EnableTelemetry bool
-
-	// BatchSize is the number of metrics to batch before flushing
-	BatchSize int
-
-	// FlushInterval is how often to flush metrics
-	FlushInterval time.Duration
-
-	// MaxRetries is the maximum number of retry attempts
-	MaxRetries int
-
-	// RetryDelay is the base delay between retries
-	RetryDelay time.Duration
-
-	// CircuitBreakerEnabled enables circuit breaker protection
-	CircuitBreakerEnabled bool
-
-	// CircuitBreakerThreshold is failures before opening circuit
-	CircuitBreakerThreshold int
-
-	// CircuitBreakerTimeout is time before retrying after open
-	CircuitBreakerTimeout time.Duration
-}
-
-// DefaultConfig returns default telemetry configuration.
-// Note: Telemetry is disabled by default and requires explicit opt-in.
-func DefaultConfig() *Config {
-	return &Config{
-		Enabled:         false, // Disabled by default, requires explicit opt-in
-		EnableTelemetry: false,
-		BatchSize:            100,
-		FlushInterval:           5 * time.Second,
-		MaxRetries:              3,
-		RetryDelay:              100 * time.Millisecond,
-		CircuitBreakerEnabled:   true,
-		CircuitBreakerThreshold: 5,
-		CircuitBreakerTimeout:   1 * time.Minute,
-	}
-}
-```
-
-### 6.2 Configuration from Connection Parameters
-
-```go
-// ParseTelemetryConfig extracts telemetry config from connection parameters.
-func ParseTelemetryConfig(params map[string]string) *Config {
-	cfg := DefaultConfig()
-
-	// Check for enableTelemetry flag (follows client > server > default priority)
-	if v, ok := params["enableTelemetry"]; ok {
-		if v == "true" || v == "1" {
-			cfg.EnableTelemetry = true
-		} else if v == "false" || v == "0" {
-			cfg.EnableTelemetry = false
-		}
-	}
-
-	if v, ok := params["telemetry_batch_size"]; ok {
-		if size, err := strconv.Atoi(v); err == nil && size > 0 {
-			cfg.BatchSize = size
-		}
-	}
-
-	if v, ok := params["telemetry_flush_interval"]; ok {
-		if duration, err := time.ParseDuration(v); err == nil {
-			cfg.FlushInterval = duration
-		}
-	}
-
-	return cfg
-}
-```
-
-### 6.3 Feature Flag Integration
-
-```go
-// checkFeatureFlag checks if telemetry is enabled server-side.
-func checkFeatureFlag(ctx context.Context, host string, httpClient *http.Client) (bool, error) {
-	endpoint := fmt.Sprintf("https://%s/api/2.0/feature-flags", host)
-
-	req, err := http.NewRequestWithContext(ctx, "GET", endpoint, nil)
-	if err != nil {
-		return false, err
-	}
-
-	// Add query parameters
-	q := req.URL.Query()
-	q.Add("flags", "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver")
-	req.URL.RawQuery = q.Encode()
-
-	resp, err := httpClient.Do(req)
-	if err != nil {
-		return false, err
-	}
-	defer resp.Body.Close()
-
-	if resp.StatusCode != http.StatusOK {
-		return false, fmt.Errorf("feature flag check failed: %d", resp.StatusCode)
-	}
-
-	var result struct {
-		Flags map[string]bool `json:"flags"`
-	}
-	if err := json.NewDecoder(resp.Body).Decode(&result); err != nil {
-		return false, err
-	}
-
-	// Parse flag response
-	flagValue := result.Flags["databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver"]
-
-	response := &featureFlagResponse{
-		Enabled:           false,
-		RolloutPercentage: 0,
-	}
-
-	// Handle both boolean and object responses for backward compatibility
-	switch v := flagValue.(type) {
-	case bool:
-		response.Enabled = v
-		if v {
-			response.RolloutPercentage = 100
-		}
-	case map[string]interface{}:
-		if enabled, ok := v["enabled"].(bool); ok {
-			response.Enabled = enabled
-		}
-		if rollout, ok := v["rollout_percentage"].(float64); ok {
-			response.RolloutPercentage = int(rollout)
-		}
-	}
-
-	return response, nil
-}
-
-// isInRollout checks if this connection is in the rollout percentage.
-// Uses consistent hashing based on workspace ID for stable rollout.
-func isInRollout(workspaceID string, rolloutPercentage int) bool {
-	if rolloutPercentage >= 100 {
-		return true
-	}
-	if rolloutPercentage <= 0 {
-		return false
-	}
-
-	// Use consistent hashing based on workspace ID
-	h := fnv.New32a()
-	h.Write([]byte(workspaceID))
-	hash := h.Sum32()
-
-	return int(hash%100) < rolloutPercentage
-}
-```
-
-#### Synchronous Fetch Behavior
-
-**Feature flag fetching is synchronous** and may block driver initialization.
-
-**Key Characteristics:**
-- 10-second HTTP timeout per request
-- Uses RetryableClient (4 retries, exponential backoff 1s-30s)
-- 15-minute cache minimizes fetch frequency
-- Thundering herd protection (only one fetch per host at a time)
-
-**When It Blocks:**
-- First connection to host: blocks for HTTP fetch (up to ~70s with retries)
-- Cache expiry (every 15 min): first caller blocks, others return stale cache
-- Concurrent callers: only first blocks, others return stale cache immediately
-
-**Why synchronous:** Simple, deterministic, 99% cache hit rate, matches JDBC driver.
-
-### 6.4 Config Overlay Pattern
-
-**UPDATE (Phase 4-5):** The telemetry system now uses a **config overlay pattern** that provides a consistent, clear priority model. This pattern is designed to be reusable across all driver configurations.
-
-#### Config Overlay Priority (highest to lowest):
-
-1. **Client Config** - Explicitly set by user (overrides server)
-2. **Server Config** - Feature flag controls when client doesn't specify
-3. **Fail-Safe Default** - Disabled when server unavailable
-
-This approach eliminates the need for special bypass flags like `forceEnableTelemetry` because client config naturally has priority.
-
-#### Implementation:
-
-```go
-// EnableTelemetry is a pointer to distinguish three states:
-// - nil: not set by client (use server feature flag)
-// - true: client wants enabled (overrides server)
-// - false: client wants disabled (overrides server)
-type Config struct {
-	EnableTelemetry *bool
-	// ... other fields
-}
-
-// isTelemetryEnabled implements config overlay
-func isTelemetryEnabled(ctx context.Context, cfg *Config, host string, httpClient *http.Client) bool {
-	// Priority 1: Client explicitly set (overrides everything)
-	if cfg.EnableTelemetry != nil {
-		return *cfg.EnableTelemetry
-	}
+## Export Format
 
-	// Priority 2: Check server-side feature flag
-	flagCache := getFeatureFlagCache()
-	serverEnabled, err := flagCache.isTelemetryEnabled(ctx, host, httpClient)
-	if err != nil {
-		// Priority 3: Fail-safe default (disabled)
-		return false
-	}
+Metrics are sent as a `TelemetryRequest` with a `protoLogs` array. Each entry is a JSON-encoded `TelemetryFrontendLog` aligned with the `OssSqlDriverTelemetryLog` proto schema:
 
-	return serverEnabled
+```json
+{
+  "uploadTime": 1234567890000,
+  "items": [],
+  "protoLogs": [
+    "{\"frontend_log_event_id\":\"20240101120000-a1b2c3d4e5f6g7h8\",\"context\":{...},\"entry\":{\"sql_driver_log\":{\"session_id\":\"...\",\"sql_statement_id\":\"...\",\"operation_latency_ms\":42,\"sql_operation\":{\"chunk_details\":{\"total_chunks_iterated\":3},...}}}}"
+  ]
 }
 ```
 
-#### Configuration Behavior Matrix:
+Event IDs are generated with `crypto/rand` + hex encoding (no modulo bias).
 
-| Client Sets | Server Returns | Result | Explanation |
-|-------------|----------------|--------|-------------|
-| `true` | `false` | **`true`** | Client overrides server |
-| `false` | `true` | **`false`** | Client overrides server |
-| `true` | error | **`true`** | Client overrides server error |
-| unset | `true` | **`true`** | Use server config |
-| unset | `false` | **`false`** | Use server config |
-| unset | error | **`false`** | Fail-safe default |
-
-#### Configuration Parameter Summary:
-
-| Parameter | Value | Behavior | Use Case |
-|-----------|-------|----------|----------|
-| `enableTelemetry=true` | Client forces enabled | Always send telemetry (overrides server) | Testing, debugging, opt-in users |
-| `enableTelemetry=false` | Client forces disabled | Never send telemetry (overrides server) | Privacy-conscious users, opt-out |
-| *(not set)* | Use server flag | Server controls via feature flag | Default behavior - Databricks-controlled rollout |
-
-#### Benefits of Config Overlay:
-
-- ✅ **Simpler**: Client > Server > Default (3 clear layers)
-- ✅ **Consistent**: Same pattern can be used for all driver configs
-- ✅ **No bypass flags**: Client config naturally has priority
-- ✅ **Reusable**: General `ConfigValue[T]` type in `internal/config/overlay.go`
-- ✅ **Type-safe**: Uses Go generics for any config type
-
-#### General Config Overlay System:
-
-A reusable config overlay system is available in `internal/config/overlay.go`:
-
-```go
-// Generic config value that supports overlay pattern
-type ConfigValue[T any] struct {
-	value *T  // nil = unset, non-nil = client set
-}
-
-// Parse from connection params
-cv := ParseBoolConfigValue(params, "enableFeature")
-
-// Resolve with overlay priority
-result := cv.ResolveWithContext(ctx, host, httpClient, serverResolver, defaultValue)
-```
-
-**Note**: A general `ConfigValue[T]` implementation is available in `internal/config/overlay.go` for extending this pattern to other driver configurations.
-
----
-
-## 7. Privacy & Compliance
-
-### 7.1 Data Privacy
-
-**Never Collected**:
-- ❌ SQL query text
-- ❌ Query results or data values
-- ❌ Table/column names
-- ❌ User identities
-
-**Always Collected**:
-- ✅ Operation latency
-- ✅ Error codes (not messages)
-- ✅ Feature flags (boolean)
-- ✅ Statement IDs (UUIDs)
-- ✅ Driver version and runtime info
-
-### 7.2 Tag Filtering
-
-```go
-// filterTagsForExport removes tags that shouldn't be sent to Databricks.
-func filterTagsForExport(metricType string, tags map[string]interface{}) map[string]interface{} {
-	filtered := make(map[string]interface{})
-
-	for key, value := range tags {
-		if shouldExportToDatabricks(metricType, key) {
-			filtered[key] = value
-		}
-		// Tags not in allowlist are silently dropped
-	}
-
-	return filtered
-}
-```
-
----
-
-## 8. Error Handling
-
-### 8.1 Error Swallowing Strategy
-
-**Core Principle**: Every telemetry error must be swallowed with minimal logging.
-
-```go
-// recoverAndLog recovers from panics and logs at trace level.
-func recoverAndLog(operation string) {
-	if r := recover(); r != nil {
-		// Use TRACE level logging (not exposed to users by default)
-		// logger.Trace().Msgf("telemetry: %s panic: %v", operation, r)
-	}
-}
-
-// Example usage
-func (i *interceptor) afterExecute(ctx context.Context, err error) {
-	defer recoverAndLog("afterExecute")
-
-	// Telemetry logic that might panic
-	// ...
-}
-```
-
-### 8.2 Error Classification
-
-```go
-// errorClassifier classifies errors as terminal or retryable.
-type errorClassifier struct{}
-
-// isTerminalError returns true if error is terminal (non-retryable).
-func isTerminalError(err error) bool {
-	if err == nil {
-		return false
-	}
-
-	// Unwrap error to check underlying type
-	var httpErr *httpError
-	if errors.As(err, &httpErr) {
-		return isTerminalHTTPStatus(httpErr.statusCode)
-	}
-
-	// Check error message patterns
-	errMsg := err.Error()
-	terminalPatterns := []string{
-		"authentication failed",
-		"unauthorized",
-		"forbidden",
-		"not found",
-		"invalid request",
-		"syntax error",
-	}
-
-	for _, pattern := range terminalPatterns {
-		if strings.Contains(strings.ToLower(errMsg), pattern) {
-			return true
-		}
-	}
-
-	return false
-}
-
-// isTerminalHTTPStatus returns true for non-retryable HTTP status codes.
-func isTerminalHTTPStatus(status int) bool {
-	return status == 400 || status == 401 || status == 403 || status == 404
-}
-```
-
----
-
-## 9. Graceful Shutdown
-
-### 9.1 Shutdown Sequence
-
-```go
-// In connection.go
-func (c *conn) Close() error {
-	// Close telemetry before closing connection
-	if c.telemetry != nil {
-		// This is non-blocking and swallows errors
-		c.telemetry.close(context.Background())
-	}
-
-	// Release per-host resources
-	if c.cfg.telemetryEnabled {
-		getClientManager().releaseClient(c.cfg.Host)
-		getFeatureFlagCache().releaseContext(c.cfg.Host)
-	}
-
-	// ... existing connection close logic ...
-
-	return nil
-}
-```
-
-### 9.2 Client Manager Shutdown
-
-The `clientManager` now includes a `shutdown()` method that provides graceful cleanup of all telemetry clients on application shutdown. This method:
-
-- Closes all active telemetry clients regardless of reference counts
-- Logs warnings for any close failures
-- Clears the clients map to prevent memory leaks
-- Returns the last error encountered (if any)
-
-```go
-// shutdown closes all telemetry clients and clears the manager.
-// Integration points will be determined in Phase 4.
-func (m *clientManager) shutdown() error {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	var lastErr error
-	for host, holder := range m.clients {
-		if err := holder.client.close(); err != nil {
-			logger.Logger.Warn().Str("host", host).Err(err).Msg("error closing telemetry client during shutdown")
-			lastErr = err
-		}
-	}
-	// Clear the map
-	m.clients = make(map[string]*clientHolder)
-	return lastErr
-}
-```
-
-**Integration Options** (to be implemented in Phase 4):
-
-1. **Public API**: Export a `Shutdown()` function for applications to call during their shutdown sequence
-2. **Driver Hook**: Integrate with `sql.DB.Close()` or driver cleanup mechanisms
-3. **Signal Handler**: Call from application signal handlers (SIGTERM, SIGINT)
-
-### 9.3 Client Shutdown
-
-```go
-// close shuts down the telemetry client gracefully.
-func (c *telemetryClient) close() error {
-	defer recoverAndLog("client.close")
-
-	// Stop background flush
-	close(c.stopCh)
-
-	// Flush pending metrics with timeout
-	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
-	defer cancel()
-
-	c.aggregator.flush(ctx)
-
-	return nil
-}
-```
+System info (OS name/version/arch, Go runtime version) is read once and cached via `sync.Once`.
 
 ---
 
-## 10. Testing Strategy
-
-### 10.1 Unit Tests
-
-```go
-// telemetry/interceptor_test.go
-func TestInterceptor_BeforeAfterExecute(t *testing.T) {
-	agg := &mockAggregator{}
-	interceptor := &interceptor{
-		aggregator: agg,
-		enabled:    true,
-	}
-
-	ctx := context.Background()
-	ctx = interceptor.beforeExecute(ctx, "stmt-123")
-
-	// Verify metric context is attached
-	mc := getMetricContext(ctx)
-	assert.NotNil(t, mc)
-	assert.Equal(t, "stmt-123", mc.statementID)
-
-	// Simulate execution
-	time.Sleep(10 * time.Millisecond)
-
-	interceptor.afterExecute(ctx, nil)
+## Configuration
 
-	// Verify metric was recorded
-	assert.Equal(t, 1, len(agg.metrics))
-	assert.True(t, agg.metrics[0].latencyMs >= 10)
-}
-
-func TestInterceptor_ErrorHandling(t *testing.T) {
-	agg := &mockAggregator{
-		shouldPanic: true,
-	}
-	interceptor := &interceptor{
-		aggregator: agg,
-		enabled:    true,
-	}
-
-	ctx := context.Background()
-	ctx = interceptor.beforeExecute(ctx, "stmt-123")
-
-	// Should not panic even if aggregator panics
-	assert.NotPanics(t, func() {
-		interceptor.afterExecute(ctx, nil)
-	})
-}
-```
-
-### 10.2 Integration Tests
-
-```go
-// telemetry/integration_test.go
-func TestTelemetry_EndToEnd(t *testing.T) {
-	if testing.Short() {
-		t.Skip("skipping integration test")
-	}
-
-	// Create mock server
-	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		if r.URL.Path == "/api/2.0/telemetry-ext" {
-			w.WriteHeader(http.StatusOK)
-		}
-	}))
-	defer server.Close()
-
-	// Create connection with telemetry
-	cfg := &config.Config{
-		Host: server.URL,
-		TelemetryEnabled: true,
-	}
-
-	conn, err := connect(cfg)
-	require.NoError(t, err)
-	defer conn.Close()
-
-	// Execute statement
-	_, err = conn.QueryContext(context.Background(), "SELECT 1")
-	require.NoError(t, err)
-
-	// Wait for flush
-	time.Sleep(6 * time.Second)
-
-	// Verify metrics were sent to server
-	// (check server mock received requests)
-}
-```
+DSN parameter: `enableTelemetry=true|false` (parsed via `strconv.ParseBool`)
 
-### 10.3 Benchmark Tests
+Priority (highest to lowest):
+1. Client DSN setting (`enableTelemetry=true/false`) — overrides server
+2. Server feature flag (`databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver`)
+3. Default: disabled
 
-```go
-// telemetry/benchmark_test.go
-func BenchmarkInterceptor_Overhead(b *testing.B) {
-	agg := &metricsAggregator{
-		// ... initialized ...
-	}
-	interceptor := &interceptor{
-		aggregator: agg,
-		enabled:    true,
-	}
-
-	ctx := context.Background()
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		ctx = interceptor.beforeExecute(ctx, "stmt-123")
-		interceptor.afterExecute(ctx, nil)
-	}
-}
-
-func BenchmarkInterceptor_Disabled(b *testing.B) {
-	agg := &metricsAggregator{
-		// ... initialized ...
-	}
-	interceptor := &interceptor{
-		aggregator: agg,
-		enabled:    false, // Disabled
-	}
-
-	ctx := context.Background()
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		ctx = interceptor.beforeExecute(ctx, "stmt-123")
-		interceptor.afterExecute(ctx, nil)
-	}
-}
-```
+Batch size, flush interval, retry count, and retry delay are also configurable.
 
 ---
 
-## 11. Implementation Checklist
-
-**Strategy**: Build infrastructure bottom-up: Circuit Breaker → Export (POST to endpoint) → Opt-In Configuration → Collection & Aggregation → Driver Integration. This allows unit testing each layer before adding metric collection.
-
-**JIRA Tickets**:
-- **PECOBLR-1143**: Phases 1-5 (Core Infrastructure → Opt-In Configuration)
-  - **PECOBLR-1381**: Phase 6 (Collection & Aggregation) - subtask
-  - **PECOBLR-1382**: Phase 7 (Driver Integration) - subtask
-
-### Phase 1: Core Infrastructure ✅ COMPLETED
-- [x] Create `telemetry` package structure
-- [x] Implement `config.go` with configuration types (basic structure)
-- [x] Implement `tags.go` with tag definitions and filtering
-- [x] Add unit tests for configuration and tags
-
-### Phase 2: Per-Host Management ✅ COMPLETED
-- [x] Implement `featureflag.go` with caching and reference counting
-- [x] Implement `manager.go` for client management
-  - [x] Thread-safe singleton pattern with per-host client holders
-  - [x] Reference counting for automatic cleanup
-  - [x] Error handling for client start failures
-  - [x] Shutdown method for graceful application shutdown
-  - [x] Comprehensive documentation on thread-safety and connection sharing
-- [x] Implement `client.go` with minimal telemetryClient stub
-  - [x] Thread-safe start() and close() methods
-  - [x] Mutex protection for state flags
-  - [x] Detailed documentation on concurrent access requirements
-- [x] Add comprehensive unit tests for all components
-  - [x] Singleton pattern verification
-  - [x] Reference counting (increment/decrement/cleanup)
-  - [x] Concurrent access tests (100+ goroutines)
-  - [x] Shutdown scenarios (empty, with active refs, multiple hosts)
-  - [x] Race detector tests passing
-
-### Phase 3: Circuit Breaker ✅ COMPLETED
-- [x] Implement `circuitbreaker.go` with state machine
-  - [x] Implement circuit breaker states (Closed, Open, Half-Open)
-  - [x] Implement circuitBreakerManager singleton per host
-  - [x] Add configurable thresholds and timeout
-  - [x] Implement execute() method with state transitions
-  - [x] Implement failure/success tracking with sliding window algorithm
-- [x] Add comprehensive unit tests
-  - [x] Test state transitions (Closed → Open → Half-Open → Closed)
-  - [x] Test failure/success counting
-  - [x] Test timeout and retry logic
-  - [x] Test per-host circuit breaker isolation
-  - [x] Test concurrent access
-
-### Phase 4: Export Infrastructure ✅ COMPLETED
-- [x] Implement `exporter.go` with retry logic
-  - [x] Implement HTTP POST to telemetry endpoint (/api/2.0/telemetry-ext)
-  - [x] Implement retry logic with exponential backoff
-  - [x] Implement tag filtering for export (shouldExportToDatabricks)
-  - [x] Integrate with circuit breaker
-  - [x] Add error swallowing
-  - [x] Implement toExportedMetric() conversion
-  - [x] Implement telemetryPayload JSON structure
-- [x] Add unit tests for export logic
-  - [x] Test HTTP request construction
-  - [x] Test retry logic (with mock HTTP responses)
-  - [x] Test circuit breaker integration
-  - [x] Test tag filtering
-  - [x] Test error swallowing
-- [x] Add integration tests with mock HTTP server
-  - [x] Test successful export
-  - [x] Test error scenarios (4xx, 5xx)
-  - [x] Test retry behavior (exponential backoff)
-  - [x] Test circuit breaker opening/closing
-  - [x] Test context cancellation
-
-### Phase 5: Opt-In Configuration Integration ✅ COMPLETED
-- [x] Implement `isTelemetryEnabled()` with priority-based logic in config.go
-  - [x] Priority 1 (client): EnableTelemetry=true → enable regardless of server flag
-  - [x] Priority 2 (client): EnableTelemetry=false → disable regardless of server flag
-  - [x] Priority 3 (server): Server feature flag controls when client preference unset
-  - [x] Priority 4 (default): Disabled if no flags set and server check fails
-- [x] Integrate feature flag cache with opt-in logic
-  - [x] Wire up isTelemetryEnabled() to call featureFlagCache.isTelemetryEnabled()
-  - [x] Implement fallback behavior on errors (return cached value or false)
-  - [x] Add proper error handling
-- [x] Add unit tests for opt-in priority logic
-  - [x] Test enableTelemetry=false (always disabled, explicit opt-out)
-  - [x] Test enableTelemetry=true with server flag enabled
-  - [x] Test enableTelemetry=true with server flag disabled
-  - [x] Test default behavior (server flag controls)
-  - [x] Test error scenarios (server unreachable, use cached value)
-- [x] Add integration tests with mock feature flag server
-  - [x] Test opt-in priority with mock server
-  - [x] Test server error handling
-  - [x] Test unreachable server scenarios
-
-### Phase 6: Collection & Aggregation (PECOBLR-1381) ✅ COMPLETED
-- [x] Implement `interceptor.go` for metric collection
-  - [x] Implement beforeExecute() and afterExecute() hooks
-  - [x] Implement context-based metric tracking with metricContext
-  - [x] Implement latency measurement (startTime, latencyMs calculation)
-  - [x] Add tag collection methods (addTag)
-  - [x] Implement error swallowing with panic recovery
-- [x] Implement `aggregator.go` for batching
-  - [x] Implement statement-level aggregation (statementMetrics)
-  - [x] Implement batch size and flush interval logic
-  - [x] Implement background flush goroutine (flushLoop)
-  - [x] Add thread-safe metric recording
-  - [x] Implement completeStatement() for final aggregation
-- [x] Implement error classification in `errors.go`
-  - [x] Implement error type classification (terminal vs retryable)
-  - [x] Implement HTTP status code classification
-  - [x] Add error pattern matching
-  - [x] Implement isTerminalError() function
-- [x] Update `client.go` to integrate aggregator
-  - [x] Wire up aggregator with exporter
-  - [x] Implement background flush timer
-  - [x] Update start() and close() methods
-- [x] Add unit tests for collection and aggregation
-  - [x] Test interceptor metric collection and latency tracking
-  - [x] Test aggregation logic
-  - [x] Test batch flushing (size-based and time-based)
-  - [x] Test error classification
-  - [x] Test client with aggregator integration
-
-### Phase 7: Driver Integration ✅ COMPLETED
-- [x] Add telemetry initialization to `connection.go`
-  - [x] Call isTelemetryEnabled() at connection open via InitializeForConnection()
-  - [x] Initialize telemetry client via clientManager.getOrCreateClient()
-  - [x] Increment feature flag cache reference count
-  - [x] Store telemetry interceptor in connection
-- [x] Add telemetry configuration to UserConfig
-  - [x] EnableTelemetry field (client > server > default priority)
-  - [x] DSN parameter parsing
-  - [x] DeepCopy support
-- [x] Add cleanup in `Close()` methods
-  - [x] Release client manager reference in connection.Close()
-  - [x] Release feature flag cache reference via ReleaseForConnection()
-  - [x] Flush pending metrics before close
-- [x] Export necessary types and methods
-  - [x] Export Interceptor type
-  - [x] Export GetInterceptor() and Close() methods
-  - [x] Create driver integration helpers
-- [x] Basic integration tests
-  - [x] Test compilation with telemetry
-  - [x] Test no breaking changes to existing tests
-  - [x] Test graceful handling when disabled
-- [x] Statement execution hooks
-  - [x] Add beforeExecute() hook to QueryContext
-  - [x] Add afterExecute() and completeStatement() hooks to QueryContext
-  - [x] Add beforeExecute() hook to ExecContext
-  - [x] Add afterExecute() and completeStatement() hooks to ExecContext
-  - [x] Use operation handle GUID as statement ID
-
-### Phase 8: Testing & Validation
-- [ ] Run benchmark tests
-  - [ ] Measure overhead when enabled
-  - [ ] Measure overhead when disabled
-  - [ ] Ensure <1% overhead when enabled
-- [ ] Perform load testing with concurrent connections
-  - [ ] Test 100+ concurrent connections
-  - [ ] Verify per-host client sharing
-  - [ ] Verify no rate limiting with per-host clients
-- [ ] Validate graceful shutdown
-  - [ ] Test reference counting cleanup
-  - [ ] Test final flush on shutdown
-  - [ ] Test shutdown method works correctly
-- [ ] Test circuit breaker behavior
-  - [ ] Test circuit opening on repeated failures
-  - [ ] Test circuit recovery after timeout
-  - [ ] Test metrics dropped when circuit open
-- [ ] Test opt-in priority logic end-to-end
-  - [ ] Verify forceEnableTelemetry works in real driver
-  - [ ] Verify enableTelemetry works in real driver
-  - [ ] Verify server flag integration works
-- [ ] Verify privacy compliance
-  - [ ] Verify no SQL queries collected
-  - [ ] Verify no PII collected
-  - [ ] Verify tag filtering works (shouldExportToDatabricks)
-
-### Phase 9: Partial Launch Preparation
-- [ ] Document `forceEnableTelemetry` and `enableTelemetry` flags
-- [ ] Create internal testing plan for Phase 1 (use forceEnableTelemetry=true)
-- [ ] Prepare beta opt-in documentation for Phase 2 (use enableTelemetry=true)
-- [ ] Set up monitoring for rollout health metrics
-- [ ] Document rollback procedures (set server flag to false)
-
-### Phase 10: Documentation
-- [ ] Document configuration options in README
-- [ ] Add examples for opt-in flags
-- [ ] Document partial launch strategy and phases
-- [ ] Document metric tags and their meanings
-- [ ] Create troubleshooting guide
-- [ ] Document architecture and design decisions
-
----
-
-## 12. References
-
-### 12.1 Go Standards
-- [context package](https://pkg.go.dev/context)
-- [database/sql/driver](https://pkg.go.dev/database/sql/driver)
-- [net/http](https://pkg.go.dev/net/http)
-- [sync package](https://pkg.go.dev/sync)
-- [Effective Go](https://go.dev/doc/effective_go)
-
-### 12.2 Existing Code References
-
-**Databricks SQL Go Driver**:
-- `connection.go`: Connection management
-- `connector.go`: Connection factory
-- `internal/config/config.go`: Configuration patterns
-- `internal/client/client.go`: HTTP client patterns
-
-**JDBC Driver** (reference implementation):
-- `TelemetryClient.java:15`: Batching and flush logic
-- `TelemetryClientFactory.java:27`: Per-host client management
-- `CircuitBreakerManager.java:25`: Circuit breaker pattern
-- `DatabricksDriverFeatureFlagsContextFactory.java:27`: Feature flag caching
-
----
-
-## Summary
-
-This **telemetry design for Go** follows idiomatic Go patterns:
+## Privacy
 
-1. **Idiomatic Go**: Uses context, interfaces, goroutines, and channels
-2. **Minimal changes**: Wraps existing driver operations with interceptors
-3. **Type safety**: Strong typing with exported/unexported types
-4. **Error handling**: All errors swallowed, uses defer/recover patterns
-5. **Concurrency**: Thread-safe using sync primitives (RWMutex, atomic)
-6. **Testing**: Comprehensive unit, integration, and benchmark tests
-7. **Standard library**: Minimal external dependencies
+Only the following fields are exported:
+- Session ID, statement ID, workspace ID
+- Operation latency, error type (no message/stack trace)
+- Chunk count, poll count
+- Driver version, OS name, Go runtime version, result format
 
-The design enables collecting valuable usage metrics while maintaining Go best practices and ensuring zero impact on driver reliability and performance.
+Tags not in this allowlist are silently dropped before export.
diff --git a/telemetry/aggregator.go b/telemetry/aggregator.go
index 15323f08..38dd636c 100644
--- a/telemetry/aggregator.go
+++ b/telemetry/aggregator.go
@@ -8,6 +8,17 @@ import (
 	"github.com/databricks/databricks-sql-go/logger"
 )
 
+const (
+	exportWorkerCount = 10   // Fixed worker pool size — matches JDBC's newFixedThreadPool(10)
+	exportQueueSize   = 1000 // Buffered queue — effectively unbounded for normal workloads
+)
+
+// exportJob holds a batch of metrics pending async export.
+type exportJob struct {
+	ctx     context.Context
+	metrics []*telemetryMetric
+}
+
 // metricsAggregator aggregates metrics by statement and batches for export.
 type metricsAggregator struct {
 	mu sync.RWMutex
@@ -21,10 +32,10 @@ type metricsAggregator struct {
 	stopCh        chan struct{}
 	flushTimer    *time.Ticker
 
-	closeOnce sync.Once
-	ctx       context.Context    // Cancellable context for in-flight exports
-	cancel    context.CancelFunc // Cancels ctx on close
-	exportSem chan struct{}      // Bounds concurrent export goroutines
+	closeOnce   sync.Once
+	ctx         context.Context // Cancellable context — cancelled on close to stop workers
+	cancel      context.CancelFunc
+	exportQueue chan exportJob // Worker queue; drop batch only when full (matches JDBC LinkedBlockingQueue)
 }
 
 // statementMetrics holds aggregated metrics for a statement.
@@ -41,7 +52,7 @@ type statementMetrics struct {
 
 // newMetricsAggregator creates a new metrics aggregator.
 func newMetricsAggregator(exporter *telemetryExporter, cfg *Config) *metricsAggregator {
-	ctx, cancel := context.WithCancel(context.Background())
+	ctx, cancel := context.WithCancel(context.Background()) //nolint:gosec // cancel stored in agg.cancel and called on close
 	agg := &metricsAggregator{
 		statements:    make(map[string]*statementMetrics),
 		batch:         make([]*telemetryMetric, 0, cfg.BatchSize),
@@ -51,7 +62,12 @@ func newMetricsAggregator(exporter *telemetryExporter, cfg *Config) *metricsAggr
 		stopCh:        make(chan struct{}),
 		ctx:           ctx,
 		cancel:        cancel,
-		exportSem:     make(chan struct{}, 8), // Bound to 8 concurrent exports
+		exportQueue:   make(chan exportJob, exportQueueSize),
+	}
+
+	// Start fixed worker pool — matches JDBC's newFixedThreadPool(10)
+	for i := 0; i < exportWorkerCount; i++ {
+		go agg.exportWorker()
 	}
 
 	// Start background flush timer
@@ -60,12 +76,27 @@ func newMetricsAggregator(exporter *telemetryExporter, cfg *Config) *metricsAggr
 	return agg
 }
 
+// exportWorker processes export jobs from the queue until the aggregator is cancelled.
+func (agg *metricsAggregator) exportWorker() {
+	for {
+		select {
+		case job, ok := <-agg.exportQueue:
+			if !ok {
+				return
+			}
+			agg.exporter.export(job.ctx, job.metrics)
+		case <-agg.ctx.Done():
+			return
+		}
+	}
+}
+
 // recordMetric records a metric for aggregation.
 func (agg *metricsAggregator) recordMetric(ctx context.Context, metric *telemetryMetric) {
 	// Swallow all errors
 	defer func() {
 		if r := recover(); r != nil {
-			logger.Debug().Msgf("telemetry: recordMetric panic: %v", r)
+			logger.Trace().Msgf("telemetry: recordMetric panic: %v", r)
 		}
 	}()
 
@@ -74,10 +105,18 @@ func (agg *metricsAggregator) recordMetric(ctx context.Context, metric *telemetr
 
 	switch metric.metricType {
 	case "connection":
-		// Emit connection events immediately: connection lifecycle events must be captured
-		// before the connection closes, as we won't have another opportunity to flush
+		// Connection events flush immediately: lifecycle events must be captured before
+		// the connection closes, as we won't have another opportunity to flush.
 		agg.batch = append(agg.batch, metric)
-		if len(agg.batch) >= agg.batchSize {
+		agg.flushUnlocked(ctx)
+
+	case "operation":
+		agg.batch = append(agg.batch, metric)
+		// Terminal operations (session/statement close) flush immediately so metrics
+		// are not lost if the connection closes before the next batch flush — matching
+		// JDBC behavior where CLOSE_STATEMENT and DELETE_SESSION trigger immediate export.
+		opType, _ := metric.tags["operation_type"].(string)
+		if isTerminalOperationType(opType) || len(agg.batch) >= agg.batchSize {
 			agg.flushUnlocked(ctx)
 		}
 
@@ -118,8 +157,7 @@ func (agg *metricsAggregator) recordMetric(ctx context.Context, metric *telemetr
 	case "error":
 		// Check if terminal error
 		if metric.errorType != "" && isTerminalError(&simpleError{msg: metric.errorType}) {
-			// Flush terminal errors immediately: terminal errors often lead to connection
-			// termination. If we wait for the next batch/timer flush, this data may be lost
+			// Flush terminal errors immediately
 			agg.batch = append(agg.batch, metric)
 			agg.flushUnlocked(ctx)
 		} else {
@@ -135,7 +173,7 @@ func (agg *metricsAggregator) recordMetric(ctx context.Context, metric *telemetr
 func (agg *metricsAggregator) completeStatement(ctx context.Context, statementID string, failed bool) {
 	defer func() {
 		if r := recover(); r != nil {
-			logger.Debug().Msgf("telemetry: completeStatement panic: %v", r)
+			logger.Trace().Msgf("telemetry: completeStatement panic: %v", r)
 		}
 	}()
 
@@ -192,52 +230,63 @@ func (agg *metricsAggregator) flushLoop() {
 	}
 }
 
-// flush flushes pending metrics to exporter.
+// flush flushes pending metrics to exporter asynchronously (fire-and-forget).
 func (agg *metricsAggregator) flush(ctx context.Context) {
 	agg.mu.Lock()
 	defer agg.mu.Unlock()
 	agg.flushUnlocked(ctx)
 }
 
-// flushUnlocked flushes without locking (caller must hold lock).
+// flushSync flushes pending metrics synchronously, blocking until the export
+// completes. Used on connection close to guarantee delivery before returning.
+func (agg *metricsAggregator) flushSync(ctx context.Context) {
+	agg.mu.Lock()
+	if len(agg.batch) == 0 {
+		agg.mu.Unlock()
+		return
+	}
+	metrics := make([]*telemetryMetric, len(agg.batch))
+	copy(metrics, agg.batch)
+	agg.batch = agg.batch[:0]
+	agg.mu.Unlock()
+
+	agg.exporter.export(ctx, metrics)
+}
+
+// flushUnlocked submits the current batch to the export worker queue (caller must hold lock).
+// Matches JDBC's executorService.submit() — drops only when the queue is full, not merely
+// when workers are busy. Workers drain the queue until the aggregator context is cancelled.
 func (agg *metricsAggregator) flushUnlocked(ctx context.Context) {
 	if len(agg.batch) == 0 {
 		return
 	}
 
-	// Copy batch and clear
 	metrics := make([]*telemetryMetric, len(agg.batch))
 	copy(metrics, agg.batch)
 	agg.batch = agg.batch[:0]
 
-	// Acquire semaphore slot; skip export if already at capacity to prevent goroutine leaks
 	select {
-	case agg.exportSem <- struct{}{}:
+	case agg.exportQueue <- exportJob{ctx: ctx, metrics: metrics}:
 	default:
-		logger.Debug().Msg("telemetry: export semaphore full, dropping metrics batch")
-		return
+		// Queue full — drop batch silently (matches JDBC's RejectedExecutionException path)
+		logger.Debug().Msg("telemetry: export queue full, dropping metrics batch")
 	}
-
-	// Export asynchronously
-	go func() {
-		defer func() {
-			<-agg.exportSem
-			if r := recover(); r != nil {
-				logger.Debug().Msgf("telemetry: async export panic: %v", r)
-			}
-		}()
-		agg.exporter.export(ctx, metrics)
-	}()
 }
 
 // close stops the aggregator and flushes pending metrics.
-// Safe to call multiple times — subsequent calls are no-ops for the stop/cancel step.
+// Safe to call multiple times — subsequent calls are no-ops (closeOnce).
+//
+// Shutdown order matters:
+//  1. Stop periodic flush (close stopCh) so no new async exports are queued.
+//  2. Synchronously flush the current batch directly (flushSync bypasses the
+//     worker queue, so it works even after workers are stopped).
+//  3. Cancel the aggregator context to stop the 10 export worker goroutines.
 func (agg *metricsAggregator) close(ctx context.Context) error {
 	agg.closeOnce.Do(func() {
-		close(agg.stopCh)
-		agg.cancel() // Cancel in-flight periodic export goroutines
+		close(agg.stopCh)  // Stop periodic flush loop
+		agg.flushSync(ctx) // Final flush — direct export, no workers needed
+		agg.cancel()       // Stop export workers after final flush
 	})
-	agg.flush(ctx)
 	return nil
 }
 
diff --git a/telemetry/benchmark_test.go b/telemetry/benchmark_test.go
new file mode 100644
index 00000000..d4687a08
--- /dev/null
+++ b/telemetry/benchmark_test.go
@@ -0,0 +1,315 @@
+// This file contains performance benchmarks and load tests for the telemetry package.
+// Run manually — NOT part of CI:
+//
+//	go test -bench=. -benchmem -run='^$' ./telemetry/...
+//
+// Benchmarks measure telemetry overhead on the hot path and validate that telemetry
+// adds negligible latency to normal driver operations.
+package telemetry
+
+import (
+	"context"
+	"fmt"
+	"net/http"
+	"net/http/httptest"
+	"sync"
+	"sync/atomic"
+	"testing"
+	"time"
+)
+
+// BenchmarkInterceptor_Overhead_Enabled measures telemetry overhead on the hot path
+// when telemetry is enabled. This is the worst-case overhead added to every statement.
+func BenchmarkInterceptor_Overhead_Enabled(b *testing.B) {
+	cfg := DefaultConfig()
+	cfg.BatchSize = 1000
+	cfg.FlushInterval = 10 * time.Minute // suppress periodic flush during bench
+
+	exporter := newTelemetryExporter("localhost", "test-version", &http.Client{}, cfg)
+	agg := newMetricsAggregator(exporter, cfg)
+	defer agg.close(context.Background()) //nolint:errcheck
+
+	interceptor := newInterceptor(agg, true)
+	ctx := context.Background()
+
+	b.ResetTimer()
+	b.ReportAllocs()
+	for i := 0; i < b.N; i++ {
+		ctx2 := interceptor.BeforeExecute(ctx, "session-bench", fmt.Sprintf("stmt-%d", i))
+		interceptor.AfterExecute(ctx2, nil)
+	}
+}
+
+// BenchmarkInterceptor_Overhead_Disabled measures the disabled-path overhead (baseline).
+// The delta between Enabled and Disabled is the pure telemetry cost.
+func BenchmarkInterceptor_Overhead_Disabled(b *testing.B) {
+	cfg := DefaultConfig()
+	exporter := newTelemetryExporter("localhost", "test-version", &http.Client{}, cfg)
+	agg := newMetricsAggregator(exporter, cfg)
+	defer agg.close(context.Background()) //nolint:errcheck
+
+	interceptor := newInterceptor(agg, false)
+	ctx := context.Background()
+
+	b.ResetTimer()
+	b.ReportAllocs()
+	for i := 0; i < b.N; i++ {
+		ctx2 := interceptor.BeforeExecute(ctx, "session-bench", fmt.Sprintf("stmt-%d", i))
+		interceptor.AfterExecute(ctx2, nil)
+	}
+}
+
+// BenchmarkAggregator_RecordMetric measures raw metric recording throughput under
+// concurrent writes. This is the critical path for every driver operation.
+func BenchmarkAggregator_RecordMetric(b *testing.B) {
+	cfg := DefaultConfig()
+	cfg.BatchSize = 10000
+	cfg.FlushInterval = 10 * time.Minute
+
+	exporter := newTelemetryExporter("localhost", "test-version", &http.Client{}, cfg)
+	agg := newMetricsAggregator(exporter, cfg)
+	defer agg.close(context.Background()) //nolint:errcheck
+
+	ctx := context.Background()
+	metric := &telemetryMetric{
+		metricType:  "operation",
+		timestamp:   time.Now(),
+		sessionID:   "bench-session",
+		statementID: "bench-stmt",
+		latencyMs:   10,
+		tags:        map[string]interface{}{"operation_type": OperationTypeExecuteStatement},
+	}
+
+	b.ResetTimer()
+	b.ReportAllocs()
+	for i := 0; i < b.N; i++ {
+		agg.recordMetric(ctx, metric)
+	}
+}
+
+// BenchmarkExporter_Export measures export throughput against a real HTTP server.
+func BenchmarkExporter_Export(b *testing.B) {
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.WriteHeader(http.StatusOK)
+	}))
+	defer server.Close()
+
+	cfg := DefaultConfig()
+	cfg.MaxRetries = 0
+	httpClient := &http.Client{Timeout: 5 * time.Second}
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
+
+	metrics := make([]*telemetryMetric, 10)
+	for i := range metrics {
+		metrics[i] = &telemetryMetric{
+			metricType: "operation",
+			timestamp:  time.Now(),
+			sessionID:  "bench-session",
+			latencyMs:  int64(i),
+			tags:       map[string]interface{}{},
+		}
+	}
+
+	ctx := context.Background()
+
+	b.ResetTimer()
+	b.ReportAllocs()
+	for i := 0; i < b.N; i++ {
+		exporter.export(ctx, metrics)
+	}
+}
+
+// BenchmarkConcurrentConnections_PerHostSharing measures per-host client sharing
+// under concurrent access. Multiple goroutines contend for the same host entry,
+// simulating many connections to one warehouse.
+func BenchmarkConcurrentConnections_PerHostSharing(b *testing.B) {
+	manager := &clientManager{
+		clients: make(map[string]*clientHolder),
+	}
+
+	host := "bench-host.databricks.com"
+	httpClient := &http.Client{}
+	cfg := DefaultConfig()
+	cfg.FlushInterval = 10 * time.Minute
+
+	b.ResetTimer()
+	b.ReportAllocs()
+	b.RunParallel(func(pb *testing.PB) {
+		for pb.Next() {
+			client := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
+			if client != nil {
+				_ = manager.releaseClient(host)
+			}
+		}
+	})
+}
+
+// BenchmarkCircuitBreaker_Execute measures the circuit breaker hot-path overhead
+// with the circuit closed (normal operation).
+func BenchmarkCircuitBreaker_Execute(b *testing.B) {
+	cb := newCircuitBreaker(defaultCircuitBreakerConfig())
+	ctx := context.Background()
+	fn := func() error { return nil }
+
+	b.ResetTimer()
+	b.ReportAllocs()
+	for i := 0; i < b.N; i++ {
+		_ = cb.execute(ctx, fn)
+	}
+}
+
+// TestLoadTesting_ConcurrentConnections verifies correctness of the aggregator
+// and client manager under high concurrency: 100 goroutines, 50 ops each.
+func TestLoadTesting_ConcurrentConnections(t *testing.T) {
+	if testing.Short() {
+		t.Skip("skipping load test in short mode")
+	}
+
+	manager := &clientManager{
+		clients: make(map[string]*clientHolder),
+	}
+
+	host := "load-test-host.databricks.com"
+	httpClient := &http.Client{}
+	cfg := DefaultConfig()
+	cfg.FlushInterval = 10 * time.Minute
+
+	const goroutines = 100
+	const opsPerGoroutine = 50
+
+	var wg sync.WaitGroup
+	var errors int64
+
+	for i := 0; i < goroutines; i++ {
+		wg.Add(1)
+		go func() {
+			defer wg.Done()
+
+			client := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
+			if client == nil {
+				atomic.AddInt64(&errors, 1)
+				return
+			}
+
+			interceptor := client.GetInterceptor(true)
+			ctx := context.Background()
+
+			for j := 0; j < opsPerGoroutine; j++ {
+				ctx2 := interceptor.BeforeExecute(ctx, "session", fmt.Sprintf("stmt-%d", j))
+				interceptor.AfterExecute(ctx2, nil)
+			}
+
+			_ = manager.releaseClient(host)
+		}()
+	}
+
+	wg.Wait()
+
+	if atomic.LoadInt64(&errors) > 0 {
+		t.Errorf("got %d errors during concurrent load test", atomic.LoadInt64(&errors))
+	}
+
+	// All connections released — client should be removed from map
+	manager.mu.RLock()
+	remaining := len(manager.clients)
+	manager.mu.RUnlock()
+	if remaining != 0 {
+		t.Errorf("expected 0 remaining clients after all releases, got %d", remaining)
+	}
+}
+
+// TestGracefulShutdown_ReferenceCountingCleanup verifies that clients are cleaned up
+// only when the last connection for a host is released (reference counting is correct).
+func TestGracefulShutdown_ReferenceCountingCleanup(t *testing.T) {
+	manager := &clientManager{
+		clients: make(map[string]*clientHolder),
+	}
+
+	hosts := []string{
+		"host-a.databricks.com",
+		"host-b.databricks.com",
+		"host-c.databricks.com",
+	}
+	httpClient := &http.Client{}
+	cfg := DefaultConfig()
+	cfg.FlushInterval = 10 * time.Minute
+
+	// Open 3 connections per host
+	for _, host := range hosts {
+		for i := 0; i < 3; i++ {
+			if client := manager.getOrCreateClient(host, "test-version", httpClient, cfg); client == nil {
+				t.Fatalf("expected client for host %s", host)
+			}
+		}
+	}
+	if len(manager.clients) != len(hosts) {
+		t.Fatalf("expected %d clients, got %d", len(hosts), len(manager.clients))
+	}
+
+	// Release 2 of 3 connections per host — clients must remain
+	for _, host := range hosts {
+		for i := 0; i < 2; i++ {
+			if err := manager.releaseClient(host); err != nil {
+				t.Errorf("unexpected error releasing client for %s: %v", host, err)
+			}
+		}
+	}
+	if len(manager.clients) != len(hosts) {
+		t.Errorf("expected clients to remain with 1 ref each, got %d clients", len(manager.clients))
+	}
+
+	// Release the last connection per host — all clients should be cleaned up
+	for _, host := range hosts {
+		if err := manager.releaseClient(host); err != nil {
+			t.Errorf("unexpected error on final release for %s: %v", host, err)
+		}
+	}
+
+	manager.mu.RLock()
+	remaining := len(manager.clients)
+	manager.mu.RUnlock()
+	if remaining != 0 {
+		t.Errorf("expected 0 clients after all releases, got %d", remaining)
+	}
+}
+
+// TestGracefulShutdown_FinalFlush verifies that metrics buffered since the last periodic
+// flush are exported synchronously when the aggregator is closed, not silently dropped.
+func TestGracefulShutdown_FinalFlush(t *testing.T) {
+	var flushed int64
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		atomic.AddInt64(&flushed, 1)
+		w.WriteHeader(http.StatusOK)
+	}))
+	defer server.Close()
+
+	cfg := DefaultConfig()
+	cfg.BatchSize = 100
+	cfg.FlushInterval = 10 * time.Minute // prevent auto-flush
+	httpClient := &http.Client{Timeout: 5 * time.Second}
+
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
+	agg := newMetricsAggregator(exporter, cfg)
+
+	ctx := context.Background()
+
+	// Record a few metrics without filling the batch (no auto-flush triggers)
+	for i := 0; i < 5; i++ {
+		agg.recordMetric(ctx, &telemetryMetric{
+			metricType: "operation",
+			timestamp:  time.Now(),
+			sessionID:  "test-session",
+			latencyMs:  int64(i),
+			tags:       map[string]interface{}{"operation_type": OperationTypeExecuteStatement},
+		})
+	}
+
+	// Close must flush remaining metrics synchronously before returning
+	if err := agg.close(ctx); err != nil {
+		t.Errorf("expected no error on close, got %v", err)
+	}
+
+	if atomic.LoadInt64(&flushed) == 0 {
+		t.Error("expected metrics to be flushed synchronously on close, but no HTTP request was made")
+	}
+}
diff --git a/telemetry/client.go b/telemetry/client.go
index 38ed6972..25c68e88 100644
--- a/telemetry/client.go
+++ b/telemetry/client.go
@@ -31,9 +31,9 @@ type telemetryClient struct {
 }
 
 // newTelemetryClient creates a new telemetry client for the given host.
-func newTelemetryClient(host string, httpClient *http.Client, cfg *Config) *telemetryClient {
+func newTelemetryClient(host string, driverVersion string, httpClient *http.Client, cfg *Config) *telemetryClient {
 	// Create exporter
-	exporter := newTelemetryExporter(host, httpClient, cfg)
+	exporter := newTelemetryExporter(host, driverVersion, httpClient, cfg)
 
 	// Create aggregator with exporter
 	aggregator := newMetricsAggregator(exporter, cfg)
diff --git a/telemetry/config.go b/telemetry/config.go
index 7bc76d00..1238333f 100644
--- a/telemetry/config.go
+++ b/telemetry/config.go
@@ -5,8 +5,6 @@ import (
 	"net/http"
 	"strconv"
 	"time"
-
-	"github.com/databricks/databricks-sql-go/internal/config"
 )
 
 // Config holds telemetry configuration.
@@ -14,12 +12,9 @@ type Config struct {
 	// Enabled controls whether telemetry is active
 	Enabled bool
 
-	// EnableTelemetry is the client-side telemetry preference.
-	// Uses config overlay pattern: client > server > default
-	// - Unset: use server feature flag (default behavior)
-	// - Set to true: client wants telemetry enabled (overrides server)
-	// - Set to false: client wants telemetry disabled (overrides server)
-	EnableTelemetry config.ConfigValue[bool]
+	// EnableTelemetry indicates user wants telemetry enabled.
+	// Follows client > server > default priority.
+	EnableTelemetry bool
 
 	// BatchSize is the number of metrics to batch before flushing
 	BatchSize int
@@ -44,12 +39,12 @@ type Config struct {
 }
 
 // DefaultConfig returns default telemetry configuration.
-// Note: Telemetry uses config overlay - controlled by server feature flags by default.
-// Clients can override by explicitly setting enableTelemetry=true/false.
+// Note: Telemetry is disabled by default. The default will remain false until
+// server-side feature flags are wired in to control the rollout.
 func DefaultConfig() *Config {
 	return &Config{
-		Enabled:                 false,                      // Will be set based on overlay logic
-		EnableTelemetry:         config.ConfigValue[bool]{}, // Unset = use server feature flag
+		Enabled:                 false,
+		EnableTelemetry:         false,
 		BatchSize:               100,
 		FlushInterval:           5 * time.Second,
 		MaxRetries:              3,
@@ -64,12 +59,11 @@ func DefaultConfig() *Config {
 func ParseTelemetryConfig(params map[string]string) *Config {
 	cfg := DefaultConfig()
 
-	// Config overlay approach: client setting overrides server feature flag
-	// Priority:
-	//   1. Client explicit setting (enableTelemetry=true/false) - overrides server
-	//   2. Server feature flag (when client doesn't set) - server controls
-	//   3. Default disabled (when server flag unavailable) - fail-safe
-	cfg.EnableTelemetry = config.ParseBoolConfigValue(params, "enableTelemetry")
+	if v, ok := params["enableTelemetry"]; ok {
+		if b, err := strconv.ParseBool(v); err == nil {
+			cfg.EnableTelemetry = b
+		}
+	}
 
 	if v, ok := params["telemetry_batch_size"]; ok {
 		if size, err := strconv.Atoi(v); err == nil && size > 0 {
@@ -87,12 +81,13 @@ func ParseTelemetryConfig(params map[string]string) *Config {
 }
 
 // isTelemetryEnabled checks if telemetry should be enabled for this connection.
-// Implements config overlay approach with clear priority order.
+// Implements the priority-based decision tree for telemetry enablement.
 //
-// Config Overlay Priority (highest to lowest):
-// 1. Client Config - enableTelemetry explicitly set (true/false) - overrides server
-// 2. Server Config - feature flag controls when client doesn't specify
-// 3. Fail-Safe Default - disabled when server flag unavailable/errors
+// Priority (highest to lowest):
+// 1. enableTelemetry=true - Client opt-in (server feature flag still consulted)
+// 2. enableTelemetry=false - Explicit opt-out (always disabled)
+// 3. Server Feature Flag Only - Default behavior (Databricks-controlled)
+// 4. Default - Disabled (false)
 //
 // Parameters:
 //   - ctx: Context for the request
@@ -102,18 +97,23 @@ func ParseTelemetryConfig(params map[string]string) *Config {
 //
 // Returns:
 //   - bool: true if telemetry should be enabled, false otherwise
-func isTelemetryEnabled(ctx context.Context, cfg *Config, host string, httpClient *http.Client) bool {
-	// Priority 1: Client explicitly set (overrides server)
-	if cfg.EnableTelemetry.IsSet() {
-		val, _ := cfg.EnableTelemetry.Get()
-		return val
+func isTelemetryEnabled(ctx context.Context, cfg *Config, host string, driverVersion string, httpClient *http.Client) bool {
+	// Priority 1 & 2: Respect client preference when explicitly set
+	// enableTelemetry=false → always disabled; enableTelemetry=true → check server flag
+	// When enableTelemetry is explicitly set to false, respect that
+	if !cfg.EnableTelemetry {
+		return false
 	}
 
-	// Priority 2: Check server-side feature flag
+	// Priority 3 & 4: Check server-side feature flag
+	// This handles both:
+	// - User explicitly opted in (enableTelemetry=true) - respect server decision
+	// - Default behavior (no explicit setting) - server controls enablement
 	flagCache := getFeatureFlagCache()
-	serverEnabled, err := flagCache.isTelemetryEnabled(ctx, host, httpClient)
+	serverEnabled, err := flagCache.isTelemetryEnabled(ctx, host, driverVersion, httpClient)
 	if err != nil {
-		// Priority 3: Fail-safe default (disabled)
+		// On error, respect default (disabled)
+		// This ensures telemetry failures don't impact driver operation
 		return false
 	}
 
diff --git a/telemetry/config_test.go b/telemetry/config_test.go
index d5ecdc2b..55f24de3 100644
--- a/telemetry/config_test.go
+++ b/telemetry/config_test.go
@@ -2,26 +2,18 @@ package telemetry
 
 import (
 	"context"
-	"encoding/json"
 	"net/http"
 	"net/http/httptest"
 	"testing"
 	"time"
-
-	"github.com/databricks/databricks-sql-go/internal/config"
 )
 
 func TestDefaultConfig(t *testing.T) {
 	cfg := DefaultConfig()
 
-	// Verify telemetry uses config overlay (nil = use server flag)
+	// Verify telemetry is disabled by default
 	if cfg.Enabled {
-		t.Error("Expected Enabled to be false by default")
-	}
-
-	// Verify EnableTelemetry is unset (config overlay - use server flag)
-	if cfg.EnableTelemetry.IsSet() {
-		t.Error("Expected EnableTelemetry to be unset (use server flag), got set")
+		t.Error("Expected telemetry to be disabled by default, got enabled")
 	}
 
 	// Verify other defaults
@@ -58,9 +50,9 @@ func TestParseTelemetryConfig_EmptyParams(t *testing.T) {
 	params := map[string]string{}
 	cfg := ParseTelemetryConfig(params)
 
-	// Should return defaults - EnableTelemetry unset means use server flag
-	if cfg.EnableTelemetry.IsSet() {
-		t.Error("Expected EnableTelemetry to be unset (use server flag) when no params provided")
+	// Should return defaults
+	if cfg.Enabled {
+		t.Error("Expected telemetry to be disabled by default")
 	}
 
 	if cfg.BatchSize != 100 {
@@ -74,8 +66,7 @@ func TestParseTelemetryConfig_EnabledTrue(t *testing.T) {
 	}
 	cfg := ParseTelemetryConfig(params)
 
-	val, ok := cfg.EnableTelemetry.Get()
-	if !ok || !val {
+	if !cfg.EnableTelemetry {
 		t.Error("Expected EnableTelemetry to be true when set to 'true'")
 	}
 }
@@ -86,8 +77,7 @@ func TestParseTelemetryConfig_Enabled1(t *testing.T) {
 	}
 	cfg := ParseTelemetryConfig(params)
 
-	val, ok := cfg.EnableTelemetry.Get()
-	if !ok || !val {
+	if !cfg.EnableTelemetry {
 		t.Error("Expected EnableTelemetry to be true when set to '1'")
 	}
 }
@@ -98,8 +88,7 @@ func TestParseTelemetryConfig_EnabledFalse(t *testing.T) {
 	}
 	cfg := ParseTelemetryConfig(params)
 
-	val, ok := cfg.EnableTelemetry.Get()
-	if !ok || val {
+	if cfg.EnableTelemetry {
 		t.Error("Expected EnableTelemetry to be false when set to 'false'")
 	}
 }
@@ -182,8 +171,7 @@ func TestParseTelemetryConfig_MultipleParams(t *testing.T) {
 	}
 	cfg := ParseTelemetryConfig(params)
 
-	val, ok := cfg.EnableTelemetry.Get()
-	if !ok || !val {
+	if !cfg.EnableTelemetry {
 		t.Error("Expected EnableTelemetry to be true")
 	}
 
@@ -201,56 +189,41 @@ func TestParseTelemetryConfig_MultipleParams(t *testing.T) {
 	}
 }
 
-// TestIsTelemetryEnabled_ClientOverrideEnabled tests Priority 1: client explicitly enables (overrides server)
-func TestIsTelemetryEnabled_ClientOverrideEnabled(t *testing.T) {
-	// Setup: Create a server that returns disabled
+// TestIsTelemetryEnabled_ExplicitOptOut tests Priority 1 (client opt-out): enableTelemetry=false
+func TestIsTelemetryEnabled_ExplicitOptOut(t *testing.T) {
+	// Setup: Create a server that returns enabled
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		// Server says disabled, but client override should win
-		resp := map[string]interface{}{
-			"flags": map[string]bool{
-				"databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver": false,
-			},
-		}
-		_ = json.NewEncoder(w).Encode(resp)
+		// Even if server says enabled, explicit opt-out should disable
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"flags": [{"name": "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver", "value": "true"}], "ttl_seconds": 300}`))
 	}))
 	defer server.Close()
 
 	cfg := &Config{
-		EnableTelemetry: config.NewConfigValue(true), // Priority 1: Client explicitly enables
+		EnableTelemetry: false, // Priority 2: Explicit opt-out
 	}
 
 	ctx := context.Background()
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
-	// Setup feature flag cache context
-	flagCache := getFeatureFlagCache()
-	flagCache.getOrCreateContext(server.URL)
-	defer flagCache.releaseContext(server.URL)
-
-	// Client override should bypass server check
-	result := isTelemetryEnabled(ctx, cfg, server.URL, httpClient)
+	result := isTelemetryEnabled(ctx, cfg, server.URL, "test-version", httpClient)
 
-	if !result {
-		t.Error("Expected telemetry to be enabled when client explicitly sets enableTelemetry=true, got disabled")
+	if result {
+		t.Error("Expected telemetry to be disabled with EnableTelemetry=false, got enabled")
 	}
 }
 
-// TestIsTelemetryEnabled_ClientOverrideDisabled tests Priority 1: client explicitly disables (overrides server)
-func TestIsTelemetryEnabled_ClientOverrideDisabled(t *testing.T) {
+// TestIsTelemetryEnabled_UserOptInServerEnabled tests Priority 1 (client opt-in): user opts in + server enabled
+func TestIsTelemetryEnabled_UserOptInServerEnabled(t *testing.T) {
 	// Setup: Create a server that returns enabled
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		// Server says enabled, but client override should win
-		resp := map[string]interface{}{
-			"flags": map[string]bool{
-				"databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver": true,
-			},
-		}
-		_ = json.NewEncoder(w).Encode(resp)
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"flags": [{"name": "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver", "value": "true"}], "ttl_seconds": 300}`))
 	}))
 	defer server.Close()
 
 	cfg := &Config{
-		EnableTelemetry: config.NewConfigValue(false), // Priority 1: Client explicitly disables
+		EnableTelemetry: true, // User wants telemetry
 	}
 
 	ctx := context.Background()
@@ -261,28 +234,24 @@ func TestIsTelemetryEnabled_ClientOverrideDisabled(t *testing.T) {
 	flagCache.getOrCreateContext(server.URL)
 	defer flagCache.releaseContext(server.URL)
 
-	result := isTelemetryEnabled(ctx, cfg, server.URL, httpClient)
+	result := isTelemetryEnabled(ctx, cfg, server.URL, "test-version", httpClient)
 
-	if result {
-		t.Error("Expected telemetry to be disabled when client explicitly sets enableTelemetry=false, got enabled")
+	if !result {
+		t.Error("Expected telemetry to be enabled when user opts in and server allows, got disabled")
 	}
 }
 
-// TestIsTelemetryEnabled_ServerEnabled tests Priority 2: server flag enables (client didn't set)
-func TestIsTelemetryEnabled_ServerEnabled(t *testing.T) {
-	// Setup: Create a server that returns enabled
+// TestIsTelemetryEnabled_UserOptInServerDisabled tests: user opts in but server disabled
+func TestIsTelemetryEnabled_UserOptInServerDisabled(t *testing.T) {
+	// Setup: Create a server that returns disabled
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		resp := map[string]interface{}{
-			"flags": map[string]bool{
-				"databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver": true,
-			},
-		}
-		_ = json.NewEncoder(w).Encode(resp)
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"flags": [{"name": "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver", "value": "false"}], "ttl_seconds": 300}`))
 	}))
 	defer server.Close()
 
 	cfg := &Config{
-		EnableTelemetry: config.ConfigValue[bool]{}, // Client didn't set - use server flag
+		EnableTelemetry: true, // User wants telemetry
 	}
 
 	ctx := context.Background()
@@ -293,28 +262,24 @@ func TestIsTelemetryEnabled_ServerEnabled(t *testing.T) {
 	flagCache.getOrCreateContext(server.URL)
 	defer flagCache.releaseContext(server.URL)
 
-	result := isTelemetryEnabled(ctx, cfg, server.URL, httpClient)
+	result := isTelemetryEnabled(ctx, cfg, server.URL, "test-version", httpClient)
 
-	if !result {
-		t.Error("Expected telemetry to be enabled when server flag is true, got disabled")
+	if result {
+		t.Error("Expected telemetry to be disabled when server disables it, got enabled")
 	}
 }
 
-// TestIsTelemetryEnabled_ServerDisabled tests Priority 2: server flag disables (client didn't set)
-func TestIsTelemetryEnabled_ServerDisabled(t *testing.T) {
-	// Setup: Create a server that returns disabled
+// TestIsTelemetryEnabled_ServerFlagOnly tests: default EnableTelemetry=false is always disabled
+func TestIsTelemetryEnabled_ServerFlagOnly(t *testing.T) {
+	// Setup: Create a server that returns enabled
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		resp := map[string]interface{}{
-			"flags": map[string]bool{
-				"databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver": false,
-			},
-		}
-		_ = json.NewEncoder(w).Encode(resp)
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"flags": [{"name": "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver", "value": "true"}], "ttl_seconds": 300}`))
 	}))
 	defer server.Close()
 
 	cfg := &Config{
-		EnableTelemetry: config.ConfigValue[bool]{}, // Client didn't set - use server flag
+		EnableTelemetry: false, // Default: no explicit user preference
 	}
 
 	ctx := context.Background()
@@ -325,29 +290,29 @@ func TestIsTelemetryEnabled_ServerDisabled(t *testing.T) {
 	flagCache.getOrCreateContext(server.URL)
 	defer flagCache.releaseContext(server.URL)
 
-	result := isTelemetryEnabled(ctx, cfg, server.URL, httpClient)
+	result := isTelemetryEnabled(ctx, cfg, server.URL, "test-version", httpClient)
 
+	// When enableTelemetry is false (default), should return false (Priority 2)
 	if result {
-		t.Error("Expected telemetry to be disabled when server flag is false, got enabled")
+		t.Error("Expected telemetry to be disabled with default EnableTelemetry=false, got enabled")
 	}
 }
 
-// TestIsTelemetryEnabled_FailSafeDefault tests Priority 3: default disabled when server unavailable
-func TestIsTelemetryEnabled_FailSafeDefault(t *testing.T) {
+// TestIsTelemetryEnabled_Default tests Priority 5: default disabled
+func TestIsTelemetryEnabled_Default(t *testing.T) {
 	cfg := DefaultConfig()
 
 	ctx := context.Background()
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
-	// No server available, should default to disabled (fail-safe)
-	result := isTelemetryEnabled(ctx, cfg, "nonexistent-host", httpClient)
+	result := isTelemetryEnabled(ctx, cfg, "test-host", "test-version", httpClient)
 
 	if result {
-		t.Error("Expected telemetry to be disabled when server unavailable (fail-safe), got enabled")
+		t.Error("Expected telemetry to be disabled by default, got enabled")
 	}
 }
 
-// TestIsTelemetryEnabled_ServerError tests Priority 3: fail-safe default on server error
+// TestIsTelemetryEnabled_ServerError tests error handling
 func TestIsTelemetryEnabled_ServerError(t *testing.T) {
 	// Setup: Create a server that returns error
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
@@ -356,7 +321,7 @@ func TestIsTelemetryEnabled_ServerError(t *testing.T) {
 	defer server.Close()
 
 	cfg := &Config{
-		EnableTelemetry: config.ConfigValue[bool]{}, // Client didn't set - should use server, but server errors
+		EnableTelemetry: true, // User wants telemetry
 	}
 
 	ctx := context.Background()
@@ -367,18 +332,18 @@ func TestIsTelemetryEnabled_ServerError(t *testing.T) {
 	flagCache.getOrCreateContext(server.URL)
 	defer flagCache.releaseContext(server.URL)
 
-	result := isTelemetryEnabled(ctx, cfg, server.URL, httpClient)
+	result := isTelemetryEnabled(ctx, cfg, server.URL, "test-version", httpClient)
 
-	// On error, should default to disabled (fail-safe)
+	// On error, should default to disabled
 	if result {
-		t.Error("Expected telemetry to be disabled on server error (fail-safe), got enabled")
+		t.Error("Expected telemetry to be disabled on server error, got enabled")
 	}
 }
 
-// TestIsTelemetryEnabled_ServerUnreachable tests Priority 3: fail-safe default on unreachable server
+// TestIsTelemetryEnabled_ServerUnreachable tests unreachable server
 func TestIsTelemetryEnabled_ServerUnreachable(t *testing.T) {
 	cfg := &Config{
-		EnableTelemetry: config.ConfigValue[bool]{}, // Client didn't set - should use server, but server unreachable
+		EnableTelemetry: true, // User wants telemetry
 	}
 
 	ctx := context.Background()
@@ -390,38 +355,10 @@ func TestIsTelemetryEnabled_ServerUnreachable(t *testing.T) {
 	flagCache.getOrCreateContext(unreachableHost)
 	defer flagCache.releaseContext(unreachableHost)
 
-	result := isTelemetryEnabled(ctx, cfg, unreachableHost, httpClient)
+	result := isTelemetryEnabled(ctx, cfg, unreachableHost, "test-version", httpClient)
 
-	// On error, should default to disabled (fail-safe)
+	// On error, should default to disabled
 	if result {
-		t.Error("Expected telemetry to be disabled when server unreachable (fail-safe), got enabled")
-	}
-}
-
-// TestIsTelemetryEnabled_ClientOverridesServerError tests Priority 1 > Priority 3
-func TestIsTelemetryEnabled_ClientOverridesServerError(t *testing.T) {
-	// Setup: Create a server that returns error
-	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		w.WriteHeader(http.StatusInternalServerError)
-	}))
-	defer server.Close()
-
-	cfg := &Config{
-		EnableTelemetry: config.NewConfigValue(true), // Client explicitly enables - should override server error
-	}
-
-	ctx := context.Background()
-	httpClient := &http.Client{Timeout: 5 * time.Second}
-
-	// Setup feature flag cache context
-	flagCache := getFeatureFlagCache()
-	flagCache.getOrCreateContext(server.URL)
-	defer flagCache.releaseContext(server.URL)
-
-	result := isTelemetryEnabled(ctx, cfg, server.URL, httpClient)
-
-	// Client override should work even when server errors
-	if !result {
-		t.Error("Expected telemetry to be enabled when client explicitly sets true, even with server error, got disabled")
+		t.Error("Expected telemetry to be disabled when server unreachable, got enabled")
 	}
 }
diff --git a/telemetry/driver_integration.go b/telemetry/driver_integration.go
index 5dcd2f71..743b4107 100644
--- a/telemetry/driver_integration.go
+++ b/telemetry/driver_integration.go
@@ -3,6 +3,7 @@ package telemetry
 import (
 	"context"
 	"net/http"
+	"time"
 
 	"github.com/databricks/databricks-sql-go/internal/config"
 )
@@ -14,6 +15,7 @@ import (
 // Parameters:
 //   - ctx: Context for the initialization
 //   - host: Databricks host
+//   - driverVersion: Driver version string
 //   - httpClient: HTTP client for making requests
 //   - enableTelemetry: Client config overlay (unset = check server flag, true/false = override server)
 //
@@ -22,30 +24,50 @@ import (
 func InitializeForConnection(
 	ctx context.Context,
 	host string,
+	driverVersion string,
 	httpClient *http.Client,
 	enableTelemetry config.ConfigValue[bool],
+	batchSize int,
+	flushInterval time.Duration,
 ) *Interceptor {
-	// Create telemetry config and apply client overlay
+	// Create telemetry config and apply client overlay.
+	// ConfigValue[bool] semantics:
+	//   - unset  → true (let server feature flag decide)
+	//   - true   → true (server feature flag still consulted)
+	//   - false  → false (explicitly disabled, skip server flag check)
 	cfg := DefaultConfig()
-	cfg.EnableTelemetry = enableTelemetry
+	if val, isSet := enableTelemetry.Get(); isSet {
+		cfg.EnableTelemetry = val
+	} else {
+		cfg.EnableTelemetry = true // Unset: default to enabled, server flag decides
+	}
+	if batchSize > 0 {
+		cfg.BatchSize = batchSize
+	}
+	if flushInterval > 0 {
+		cfg.FlushInterval = flushInterval
+	}
+
+	// Get feature flag cache context FIRST (for reference counting)
+	flagCache := getFeatureFlagCache()
+	flagCache.getOrCreateContext(host)
 
 	// Check if telemetry should be enabled
-	if !isTelemetryEnabled(ctx, cfg, host, httpClient) {
+	enabled := isTelemetryEnabled(ctx, cfg, host, driverVersion, httpClient)
+	if !enabled {
+		flagCache.releaseContext(host)
 		return nil
 	}
 
 	// Get or create telemetry client for this host
 	clientMgr := getClientManager()
-	telemetryClient := clientMgr.getOrCreateClient(host, httpClient, cfg)
+	telemetryClient := clientMgr.getOrCreateClient(host, driverVersion, httpClient, cfg)
 	if telemetryClient == nil {
+		// Client failed to start; release the flag cache ref we incremented above
+		flagCache.releaseContext(host)
 		return nil
 	}
 
-	// Get feature flag cache context (for reference counting)
-	flagCache := getFeatureFlagCache()
-	flagCache.getOrCreateContext(host)
-
-	// Return interceptor
 	return telemetryClient.GetInterceptor(true)
 }
 
diff --git a/telemetry/errors.go b/telemetry/errors.go
index edc2319e..0702dd06 100644
--- a/telemetry/errors.go
+++ b/telemetry/errors.go
@@ -78,14 +78,6 @@ func classifyError(err error) string {
 	return "error"
 }
 
-// isRetryableError returns true if the error is retryable.
-// This is the inverse of isTerminalError.
-//
-//nolint:deadcode,unused // Will be used in Phase 8+
-func isRetryableError(err error) bool {
-	return !isTerminalError(err)
-}
-
 // httpError represents an HTTP error with status code.
 
 type httpError struct {
@@ -97,16 +89,6 @@ func (e *httpError) Error() string {
 	return e.message
 }
 
-// newHTTPError creates a new HTTP error.
-//
-//nolint:deadcode,unused // Will be used in Phase 8+
-func newHTTPError(statusCode int, message string) error {
-	return &httpError{
-		statusCode: statusCode,
-		message:    message,
-	}
-}
-
 // isTerminalHTTPStatus returns true for non-retryable HTTP status codes.
 
 func isTerminalHTTPStatus(status int) bool {
diff --git a/telemetry/exporter.go b/telemetry/exporter.go
index 307fb85a..53fbc14a 100644
--- a/telemetry/exporter.go
+++ b/telemetry/exporter.go
@@ -5,6 +5,7 @@ import (
 	"context"
 	"encoding/json"
 	"fmt"
+	"io"
 	"net/http"
 	"strings"
 	"time"
@@ -12,9 +13,17 @@ import (
 	"github.com/databricks/databricks-sql-go/logger"
 )
 
+const (
+	telemetryEndpointPath = "/telemetry-ext"
+	httpPrefix            = "http://"
+	httpsPrefix           = "https://"
+	defaultScheme         = "https://"
+)
+
 // telemetryExporter exports metrics to Databricks telemetry service.
 type telemetryExporter struct {
 	host           string
+	driverVersion  string
 	httpClient     *http.Client
 	circuitBreaker *circuitBreaker
 	cfg            *Config
@@ -32,11 +41,6 @@ type telemetryMetric struct {
 	tags        map[string]interface{}
 }
 
-// telemetryPayload is the JSON structure sent to Databricks.
-type telemetryPayload struct {
-	Metrics []*exportedMetric `json:"metrics"`
-}
-
 // exportedMetric is a single metric in the payload.
 type exportedMetric struct {
 	MetricType  string                 `json:"metric_type"`
@@ -49,10 +53,19 @@ type exportedMetric struct {
 	Tags        map[string]interface{} `json:"tags,omitempty"`
 }
 
+// ensureHTTPScheme adds https:// prefix to host if no scheme is present.
+func ensureHTTPScheme(host string) string {
+	if strings.HasPrefix(host, httpPrefix) || strings.HasPrefix(host, httpsPrefix) {
+		return host
+	}
+	return defaultScheme + host
+}
+
 // newTelemetryExporter creates a new exporter.
-func newTelemetryExporter(host string, httpClient *http.Client, cfg *Config) *telemetryExporter {
+func newTelemetryExporter(host string, driverVersion string, httpClient *http.Client, cfg *Config) *telemetryExporter {
 	return &telemetryExporter{
 		host:           host,
+		driverVersion:  driverVersion,
 		httpClient:     httpClient,
 		circuitBreaker: getCircuitBreakerManager().getCircuitBreaker(host),
 		cfg:            cfg,
@@ -65,8 +78,7 @@ func (e *telemetryExporter) export(ctx context.Context, metrics []*telemetryMetr
 	// Swallow all errors and panics
 	defer func() {
 		if r := recover(); r != nil {
-			// Intentionally swallow panic - telemetry must not impact driver
-			logger.Debug().Msgf("telemetry: export panic: %v", r)
+			logger.Trace().Msgf("telemetry: export panic: %v", r)
 		}
 	}()
 
@@ -81,38 +93,27 @@ func (e *telemetryExporter) export(ctx context.Context, metrics []*telemetryMetr
 	}
 
 	if err != nil {
-		// Intentionally swallow error - telemetry must not impact driver
-		_ = err // Log at trace level only: logger.Trace().Msgf("telemetry: export error: %v", err)
+		logger.Trace().Msgf("telemetry: export error: %v", err)
 	}
 }
 
 // doExport performs the actual export with retries and exponential backoff.
 func (e *telemetryExporter) doExport(ctx context.Context, metrics []*telemetryMetric) error {
-	// Convert metrics to exported format with tag filtering
-	exportedMetrics := make([]*exportedMetric, 0, len(metrics))
-	for _, m := range metrics {
-		exportedMetrics = append(exportedMetrics, m.toExportedMetric())
-	}
-
-	// Create payload
-	payload := &telemetryPayload{
-		Metrics: exportedMetrics,
+	// Create telemetry request with base64-encoded logs
+	request, err := createTelemetryRequest(metrics, e.driverVersion)
+	if err != nil {
+		return fmt.Errorf("failed to create telemetry request: %w", err)
 	}
 
-	// Serialize metrics
-	data, err := json.Marshal(payload)
+	// Serialize request
+	data, err := json.Marshal(request)
 	if err != nil {
-		return fmt.Errorf("failed to marshal metrics: %w", err)
+		return fmt.Errorf("failed to marshal request: %w", err)
 	}
 
 	// Determine endpoint
-	// Support both plain hosts and full URLs (for testing)
-	var endpoint string
-	if strings.HasPrefix(e.host, "http://") || strings.HasPrefix(e.host, "https://") {
-		endpoint = fmt.Sprintf("%s/telemetry-ext", e.host)
-	} else {
-		endpoint = fmt.Sprintf("https://%s/telemetry-ext", e.host)
-	}
+	hostURL := ensureHTTPScheme(e.host)
+	endpoint := hostURL + telemetryEndpointPath
 
 	// Retry logic with exponential backoff
 	maxRetries := e.cfg.MaxRetries
@@ -122,7 +123,6 @@ func (e *telemetryExporter) doExport(ctx context.Context, metrics []*telemetryMe
 			backoff := time.Duration(1<<uint(attempt-1)) * e.cfg.RetryDelay
 			select {
 			case <-time.After(backoff):
-				// Backoff completed, continue to retry
 			case <-ctx.Done():
 				return ctx.Err()
 			}
@@ -145,8 +145,9 @@ func (e *telemetryExporter) doExport(ctx context.Context, metrics []*telemetryMe
 			continue
 		}
 
-		// Close body to allow connection reuse
-		resp.Body.Close()
+		// Read response body
+		_, _ = io.ReadAll(resp.Body)
+		resp.Body.Close() //nolint:errcheck,gosec // G104: close after response is read
 
 		// Check status code
 		if resp.StatusCode >= 200 && resp.StatusCode < 300 {
diff --git a/telemetry/exporter_test.go b/telemetry/exporter_test.go
index bb772453..864f0c7e 100644
--- a/telemetry/exporter_test.go
+++ b/telemetry/exporter_test.go
@@ -16,7 +16,7 @@ func TestNewTelemetryExporter(t *testing.T) {
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 	host := "test-host"
 
-	exporter := newTelemetryExporter(host, httpClient, cfg)
+	exporter := newTelemetryExporter(host, "test-version", httpClient, cfg)
 
 	if exporter.host != host {
 		t.Errorf("Expected host %s, got %s", host, exporter.host)
@@ -54,15 +54,15 @@ func TestExport_Success(t *testing.T) {
 			t.Errorf("Expected Content-Type application/json, got %s", r.Header.Get("Content-Type"))
 		}
 
-		// Verify payload structure
+		// Verify payload structure (new TelemetryRequest format)
 		body, _ := io.ReadAll(r.Body)
-		var payload telemetryPayload
+		var payload TelemetryRequest
 		if err := json.Unmarshal(body, &payload); err != nil {
 			t.Errorf("Failed to unmarshal payload: %v", err)
 		}
 
-		if len(payload.Metrics) != 1 {
-			t.Errorf("Expected 1 metric, got %d", len(payload.Metrics))
+		if len(payload.ProtoLogs) != 1 {
+			t.Errorf("Expected 1 proto log, got %d", len(payload.ProtoLogs))
 		}
 
 		w.WriteHeader(http.StatusOK)
@@ -73,7 +73,7 @@ func TestExport_Success(t *testing.T) {
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
 	// Use full server URL for testing
-	exporter := newTelemetryExporter(server.URL, httpClient, cfg)
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
 
 	metrics := []*telemetryMetric{
 		{
@@ -113,7 +113,7 @@ func TestExport_RetryOn5xx(t *testing.T) {
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
 	// Use full server URL for testing
-	exporter := newTelemetryExporter(server.URL, httpClient, cfg)
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
 
 	metrics := []*telemetryMetric{
 		{
@@ -145,7 +145,7 @@ func TestExport_NonRetryable4xx(t *testing.T) {
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
 	// Use full server URL for testing
-	exporter := newTelemetryExporter(server.URL, httpClient, cfg)
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
 
 	metrics := []*telemetryMetric{
 		{
@@ -181,7 +181,7 @@ func TestExport_Retry429(t *testing.T) {
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
 	// Use full server URL for testing
-	exporter := newTelemetryExporter(server.URL, httpClient, cfg)
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
 
 	metrics := []*telemetryMetric{
 		{
@@ -211,7 +211,7 @@ func TestExport_CircuitBreakerOpen(t *testing.T) {
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
 	// Use full server URL for testing
-	exporter := newTelemetryExporter(server.URL, httpClient, cfg)
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
 
 	// Open the circuit breaker by recording failures
 	cb := exporter.circuitBreaker
@@ -334,7 +334,7 @@ func TestExport_ErrorSwallowing(t *testing.T) {
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
 	// Use full server URL for testing
-	exporter := newTelemetryExporter(server.URL, httpClient, cfg)
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
 
 	metrics := []*telemetryMetric{
 		{
@@ -370,7 +370,7 @@ func TestExport_ContextCancellation(t *testing.T) {
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
 	// Use full server URL for testing
-	exporter := newTelemetryExporter(server.URL, httpClient, cfg)
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
 
 	metrics := []*telemetryMetric{
 		{
@@ -403,7 +403,7 @@ func TestExport_ExponentialBackoff(t *testing.T) {
 	httpClient := &http.Client{Timeout: 5 * time.Second}
 
 	// Use full server URL for testing
-	exporter := newTelemetryExporter(server.URL, httpClient, cfg)
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
 
 	metrics := []*telemetryMetric{
 		{
diff --git a/telemetry/featureflag.go b/telemetry/featureflag.go
index 6943e455..81696baa 100644
--- a/telemetry/featureflag.go
+++ b/telemetry/featureflag.go
@@ -6,7 +6,6 @@ import (
 	"fmt"
 	"io"
 	"net/http"
-	"strings"
 	"sync"
 	"time"
 )
@@ -16,12 +15,10 @@ const (
 	featureFlagCacheDuration = 15 * time.Minute
 	// featureFlagHTTPTimeout is the default timeout for feature flag HTTP requests
 	featureFlagHTTPTimeout = 10 * time.Second
-
-	// Feature flag names
-	// flagEnableTelemetry controls whether telemetry is enabled for the Go driver
-	flagEnableTelemetry = "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver"
-	// Add more feature flags here as needed:
-	// flagEnableNewFeature = "databricks.partnerplatform.clientConfigsFeatureFlags.enableNewFeatureForGoDriver"
+	// featureFlagEndpointPath is the path for feature flag endpoint
+	featureFlagEndpointPath = "/api/2.0/connector-service/feature-flags/GOLANG/"
+	// featureFlagName is the name of the Go driver telemetry feature flag
+	featureFlagName = "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver"
 )
 
 // featureFlagCache manages feature flag state per host with reference counting.
@@ -33,12 +30,12 @@ type featureFlagCache struct {
 
 // featureFlagContext holds feature flag state and reference count for a host.
 type featureFlagContext struct {
-	mu            sync.RWMutex    // protects flags, lastFetched, fetching
-	flags         map[string]bool // cached feature flags by name
-	lastFetched   time.Time       // when flags were last fetched
-	refCount      int             // protected by featureFlagCache.mu
-	cacheDuration time.Duration   // how long to cache flags
-	fetching      bool            // true if a fetch is in progress
+	mu            sync.RWMutex // protects enabled, lastFetched, fetching
+	enabled       *bool
+	lastFetched   time.Time
+	refCount      int // protected by featureFlagCache.mu
+	cacheDuration time.Duration
+	fetching      bool // true if a fetch is in progress
 }
 
 var (
@@ -87,125 +84,89 @@ func (c *featureFlagCache) releaseContext(host string) {
 	}
 }
 
-// getFeatureFlag retrieves a specific feature flag value for the host.
-// This is the generic method that handles caching and fetching for any flag.
+// isTelemetryEnabled checks if telemetry is enabled for the host.
 // Uses cached value if available and not expired.
-func (c *featureFlagCache) getFeatureFlag(ctx context.Context, host string, httpClient *http.Client, flagName string) (bool, error) {
+func (c *featureFlagCache) isTelemetryEnabled(ctx context.Context, host string, driverVersion string, httpClient *http.Client) (bool, error) {
 	c.mu.RLock()
 	flagCtx, exists := c.contexts[host]
 	c.mu.RUnlock()
 
-	// If context doesn't exist, create it and make initial blocking fetch
 	if !exists {
-		c.mu.Lock()
-		// Double-check after acquiring write lock
-		flagCtx, exists = c.contexts[host]
-		if !exists {
-			flagCtx = &featureFlagContext{
-				cacheDuration: featureFlagCacheDuration,
-				fetching:      true, // Mark as fetching
-			}
-			c.contexts[host] = flagCtx
-		}
-		c.mu.Unlock()
-
-		// If we just created the context, make the initial blocking fetch
-		if !exists {
-			flags, err := fetchFeatureFlags(ctx, host, httpClient)
-
-			flagCtx.mu.Lock()
-			flagCtx.fetching = false
-			if err == nil {
-				flagCtx.flags = flags
-				flagCtx.lastFetched = time.Now()
-				result := flags[flagName]
-				flagCtx.mu.Unlock()
-				return result, nil
-			}
-			// On error for first fetch, fail-safe: return false (telemetry disabled)
-			flagCtx.mu.Unlock()
-			return false, nil
-		}
+		return false, nil
 	}
 
-	// Check if cache is valid (with proper locking)
+	// Fast path: check cache under read lock.
 	flagCtx.mu.RLock()
-	if flagCtx.flags != nil && time.Since(flagCtx.lastFetched) < flagCtx.cacheDuration {
-		// Cache is valid, return the cached flag value
-		enabled := flagCtx.flags[flagName] // returns false if flag not found
+	if flagCtx.enabled != nil && time.Since(flagCtx.lastFetched) < flagCtx.cacheDuration {
+		enabled := *flagCtx.enabled
 		flagCtx.mu.RUnlock()
 		return enabled, nil
 	}
-
-	// Check if another goroutine is already fetching
 	if flagCtx.fetching {
-		// Return cached value if available, otherwise wait
-		if flagCtx.flags != nil {
-			enabled := flagCtx.flags[flagName]
+		if flagCtx.enabled != nil {
+			enabled := *flagCtx.enabled
 			flagCtx.mu.RUnlock()
 			return enabled, nil
 		}
 		flagCtx.mu.RUnlock()
-		// No cached value and fetch in progress, return false
 		return false, nil
 	}
+	flagCtx.mu.RUnlock()
 
-	// Mark as fetching
+	// Slow path: need a write lock to set fetching=true.
+	// Re-check all conditions under write lock (double-checked locking) to avoid
+	// a data race and to prevent duplicate fetches from concurrent goroutines.
+	flagCtx.mu.Lock()
+	if flagCtx.enabled != nil && time.Since(flagCtx.lastFetched) < flagCtx.cacheDuration {
+		enabled := *flagCtx.enabled
+		flagCtx.mu.Unlock()
+		return enabled, nil
+	}
+	if flagCtx.fetching {
+		if flagCtx.enabled != nil {
+			enabled := *flagCtx.enabled
+			flagCtx.mu.Unlock()
+			return enabled, nil
+		}
+		flagCtx.mu.Unlock()
+		return false, nil
+	}
 	flagCtx.fetching = true
-	flagCtx.mu.RUnlock()
+	flagCtx.mu.Unlock()
 
-	// Fetch fresh values for all flags
-	flags, err := fetchFeatureFlags(ctx, host, httpClient)
+	// Fetch fresh value (outside lock so other readers are not blocked).
+	enabled, err := fetchFeatureFlag(ctx, host, driverVersion, httpClient)
 
-	// Update cache (with proper locking)
+	// Update cache.
 	flagCtx.mu.Lock()
 	flagCtx.fetching = false
 	if err == nil {
-		flagCtx.flags = flags
+		flagCtx.enabled = &enabled
 		flagCtx.lastFetched = time.Now()
 	}
-	// On error, keep the old cached values if they exist
 	result := false
 	var returnErr error
 	if err != nil {
-		if flagCtx.flags != nil {
-			result = flagCtx.flags[flagName]
-			returnErr = nil // Return cached value without error
+		if flagCtx.enabled != nil {
+			result = *flagCtx.enabled // Return stale cached value on error
 		} else {
 			returnErr = err
 		}
 	} else {
-		result = flags[flagName]
+		result = enabled
 	}
 	flagCtx.mu.Unlock()
 
 	return result, returnErr
 }
 
-// isTelemetryEnabled checks if telemetry is enabled for the host.
-// Uses cached value if available and not expired.
-func (c *featureFlagCache) isTelemetryEnabled(ctx context.Context, host string, httpClient *http.Client) (bool, error) {
-	return c.getFeatureFlag(ctx, host, httpClient, flagEnableTelemetry)
-}
-
 // isExpired returns true if the cache has expired.
 func (c *featureFlagContext) isExpired() bool {
-	return c.flags == nil || time.Since(c.lastFetched) > c.cacheDuration
-}
-
-// getAllFeatureFlags returns a list of all feature flags to fetch.
-// Add new flags here when adding new features.
-func getAllFeatureFlags() []string {
-	return []string{
-		flagEnableTelemetry,
-		// Add more flags here as needed:
-		// flagEnableNewFeature,
-	}
+	return c.enabled == nil || time.Since(c.lastFetched) > c.cacheDuration
 }
 
-// fetchFeatureFlags fetches multiple feature flag values from Databricks in a single request.
-// Returns a map of flag names to their boolean values.
-func fetchFeatureFlags(ctx context.Context, host string, httpClient *http.Client) (map[string]bool, error) {
+// fetchFeatureFlag fetches the feature flag value from Databricks.
+func fetchFeatureFlag(ctx context.Context, host string, driverVersion string, httpClient *http.Client) (bool, error) {
 	// Add timeout to context if it doesn't have a deadline
 	if _, hasDeadline := ctx.Deadline(); !hasDeadline {
 		var cancel context.CancelFunc
@@ -213,50 +174,50 @@ func fetchFeatureFlags(ctx context.Context, host string, httpClient *http.Client
 		defer cancel()
 	}
 
-	// Construct endpoint URL, adding https:// if not already present
-	var endpoint string
-	if strings.HasPrefix(host, "http://") || strings.HasPrefix(host, "https://") {
-		endpoint = fmt.Sprintf("%s/api/2.0/feature-flags", host)
-	} else {
-		endpoint = fmt.Sprintf("https://%s/api/2.0/feature-flags", host)
-	}
+	// Construct endpoint URL using connector-service endpoint like JDBC
+	hostURL := ensureHTTPScheme(host)
+	endpoint := fmt.Sprintf("%s%s%s", hostURL, featureFlagEndpointPath, driverVersion)
 
 	req, err := http.NewRequestWithContext(ctx, "GET", endpoint, nil)
 	if err != nil {
-		return nil, fmt.Errorf("failed to create feature flag request: %w", err)
+		return false, fmt.Errorf("failed to create feature flag request: %w", err)
 	}
 
-	// Add query parameter with comma-separated list of feature flags
-	// This fetches all flags in a single request for efficiency
-	allFlags := getAllFeatureFlags()
-	q := req.URL.Query()
-	q.Add("flags", strings.Join(allFlags, ","))
-	req.URL.RawQuery = q.Encode()
-
 	resp, err := httpClient.Do(req)
 	if err != nil {
-		return nil, fmt.Errorf("failed to fetch feature flags: %w", err)
+		return false, fmt.Errorf("failed to fetch feature flag: %w", err)
 	}
-	defer resp.Body.Close()
+	defer resp.Body.Close() //nolint:errcheck
 
 	if resp.StatusCode != http.StatusOK {
 		// Read and discard body to allow HTTP connection reuse
 		_, _ = io.Copy(io.Discard, resp.Body)
-		return nil, fmt.Errorf("feature flag check failed: %d", resp.StatusCode)
+		return false, fmt.Errorf("feature flag check failed: %d", resp.StatusCode)
+	}
+
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return false, fmt.Errorf("failed to read feature flag response: %w", err)
 	}
 
 	var result struct {
-		Flags map[string]bool `json:"flags"`
+		Flags []struct {
+			Name  string `json:"name"`
+			Value string `json:"value"`
+		} `json:"flags"`
+		TTLSeconds int `json:"ttl_seconds"`
 	}
-	if err := json.NewDecoder(resp.Body).Decode(&result); err != nil {
-		return nil, fmt.Errorf("failed to decode feature flag response: %w", err)
+	if err := json.Unmarshal(body, &result); err != nil {
+		return false, fmt.Errorf("failed to decode feature flag response: %w", err)
 	}
 
-	// Return the full map of flags
-	// Flags not present in the response will have false value when accessed
-	if result.Flags == nil {
-		return make(map[string]bool), nil
+	// Look for Go driver telemetry feature flag
+	for _, flag := range result.Flags {
+		if flag.Name == featureFlagName {
+			enabled := flag.Value == "true"
+			return enabled, nil
+		}
 	}
 
-	return result.Flags, nil
+	return false, nil
 }
diff --git a/telemetry/featureflag_test.go b/telemetry/featureflag_test.go
index b0aa519a..4ffbf07b 100644
--- a/telemetry/featureflag_test.go
+++ b/telemetry/featureflag_test.go
@@ -94,13 +94,12 @@ func TestFeatureFlagCache_IsTelemetryEnabled_Cached(t *testing.T) {
 	ctx := cache.getOrCreateContext(host)
 
 	// Set cached value
-	ctx.flags = map[string]bool{
-		flagEnableTelemetry: true,
-	}
+	enabled := true
+	ctx.enabled = &enabled
 	ctx.lastFetched = time.Now()
 
 	// Should return cached value without HTTP call
-	result, err := cache.isTelemetryEnabled(context.Background(), host, nil)
+	result, err := cache.isTelemetryEnabled(context.Background(), host, "test-version", nil)
 	if err != nil {
 		t.Errorf("Expected no error, got %v", err)
 	}
@@ -116,7 +115,7 @@ func TestFeatureFlagCache_IsTelemetryEnabled_Expired(t *testing.T) {
 		callCount++
 		w.Header().Set("Content-Type", "application/json")
 		w.WriteHeader(http.StatusOK)
-		_, _ = w.Write([]byte(`{"flags": {"databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver": true}}`))
+		_, _ = w.Write([]byte(`{"flags": [{"name": "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver", "value": "true"}], "ttl_seconds": 300}`))
 	}))
 	defer server.Close()
 
@@ -128,14 +127,13 @@ func TestFeatureFlagCache_IsTelemetryEnabled_Expired(t *testing.T) {
 	ctx := cache.getOrCreateContext(host)
 
 	// Set expired cached value
-	ctx.flags = map[string]bool{
-		flagEnableTelemetry: false,
-	}
+	enabled := false
+	ctx.enabled = &enabled
 	ctx.lastFetched = time.Now().Add(-20 * time.Minute) // Expired
 
 	// Should fetch fresh value
 	httpClient := &http.Client{}
-	result, err := cache.isTelemetryEnabled(context.Background(), host, httpClient)
+	result, err := cache.isTelemetryEnabled(context.Background(), host, "test-version", httpClient)
 	if err != nil {
 		t.Errorf("Expected no error, got %v", err)
 	}
@@ -147,7 +145,7 @@ func TestFeatureFlagCache_IsTelemetryEnabled_Expired(t *testing.T) {
 	}
 
 	// Verify cache was updated
-	if ctx.flags[flagEnableTelemetry] != true {
+	if *ctx.enabled != true {
 		t.Error("Expected cache to be updated with new value")
 	}
 }
@@ -159,15 +157,14 @@ func TestFeatureFlagCache_IsTelemetryEnabled_NoContext(t *testing.T) {
 
 	host := "non-existent-host.databricks.com"
 
-	// Should return false for non-existent context (network error expected)
-	httpClient := &http.Client{Timeout: 1 * time.Second}
-	result, err := cache.isTelemetryEnabled(context.Background(), host, httpClient)
-	// Error expected due to network failure, but should not panic
+	// Should return false for non-existent context
+	result, err := cache.isTelemetryEnabled(context.Background(), host, "test-version", nil)
+	if err != nil {
+		t.Errorf("Expected no error, got %v", err)
+	}
 	if result != false {
 		t.Error("Expected false for non-existent context")
 	}
-	// err is expected to be non-nil due to DNS/network failure, but that's okay
-	_ = err
 }
 
 func TestFeatureFlagCache_IsTelemetryEnabled_ErrorFallback(t *testing.T) {
@@ -185,14 +182,13 @@ func TestFeatureFlagCache_IsTelemetryEnabled_ErrorFallback(t *testing.T) {
 	ctx := cache.getOrCreateContext(host)
 
 	// Set cached value
-	ctx.flags = map[string]bool{
-		flagEnableTelemetry: true,
-	}
+	enabled := true
+	ctx.enabled = &enabled
 	ctx.lastFetched = time.Now().Add(-20 * time.Minute) // Expired
 
 	// Should return cached value on error
 	httpClient := &http.Client{}
-	result, err := cache.isTelemetryEnabled(context.Background(), host, httpClient)
+	result, err := cache.isTelemetryEnabled(context.Background(), host, "test-version", httpClient)
 	if err != nil {
 		t.Errorf("Expected no error (fallback to cache), got %v", err)
 	}
@@ -217,7 +213,7 @@ func TestFeatureFlagCache_IsTelemetryEnabled_ErrorNoCache(t *testing.T) {
 
 	// No cached value, should return error
 	httpClient := &http.Client{}
-	result, err := cache.isTelemetryEnabled(context.Background(), host, httpClient)
+	result, err := cache.isTelemetryEnabled(context.Background(), host, "test-version", httpClient)
 	if err == nil {
 		t.Error("Expected error when no cache available and fetch fails")
 	}
@@ -275,28 +271,28 @@ func TestFeatureFlagCache_ConcurrentAccess(t *testing.T) {
 func TestFeatureFlagContext_IsExpired(t *testing.T) {
 	tests := []struct {
 		name     string
-		flags    map[string]bool
+		enabled  *bool
 		fetched  time.Time
 		duration time.Duration
 		want     bool
 	}{
 		{
 			name:     "no cache",
-			flags:    nil,
+			enabled:  nil,
 			fetched:  time.Time{},
 			duration: 15 * time.Minute,
 			want:     true,
 		},
 		{
 			name:     "fresh cache",
-			flags:    map[string]bool{flagEnableTelemetry: true},
+			enabled:  boolPtr(true),
 			fetched:  time.Now(),
 			duration: 15 * time.Minute,
 			want:     false,
 		},
 		{
 			name:     "expired cache",
-			flags:    map[string]bool{flagEnableTelemetry: true},
+			enabled:  boolPtr(true),
 			fetched:  time.Now().Add(-20 * time.Minute),
 			duration: 15 * time.Minute,
 			want:     true,
@@ -306,7 +302,7 @@ func TestFeatureFlagContext_IsExpired(t *testing.T) {
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
 			ctx := &featureFlagContext{
-				flags:         tt.flags,
+				enabled:       tt.enabled,
 				lastFetched:   tt.fetched,
 				cacheDuration: tt.duration,
 			}
@@ -317,82 +313,73 @@ func TestFeatureFlagContext_IsExpired(t *testing.T) {
 	}
 }
 
-func TestFetchFeatureFlags_Success(t *testing.T) {
+func TestFetchFeatureFlag_Success(t *testing.T) {
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		// Verify request
+		// Verify request method
 		if r.Method != "GET" {
 			t.Errorf("Expected GET request, got %s", r.Method)
 		}
-		if r.URL.Path != "/api/2.0/feature-flags" {
-			t.Errorf("Expected /api/2.0/feature-flags path, got %s", r.URL.Path)
-		}
-
-		flags := r.URL.Query().Get("flags")
-		expectedFlag := "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver"
-		if flags != expectedFlag {
-			t.Errorf("Expected flag query param %s, got %s", expectedFlag, flags)
-		}
 
-		// Return success response
+		// Return success response using new connector-service format
 		w.Header().Set("Content-Type", "application/json")
 		w.WriteHeader(http.StatusOK)
-		_, _ = w.Write([]byte(`{"flags": {"databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver": true}}`))
+		_, _ = w.Write([]byte(`{"flags": [{"name": "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver", "value": "true"}], "ttl_seconds": 300}`))
 	}))
 	defer server.Close()
 
 	host := server.URL // Use full URL for testing
 	httpClient := &http.Client{}
 
-	flags, err := fetchFeatureFlags(context.Background(), host, httpClient)
+	enabled, err := fetchFeatureFlag(context.Background(), host, "test-version", httpClient)
 	if err != nil {
 		t.Errorf("Expected no error, got %v", err)
 	}
-	if !flags[flagEnableTelemetry] {
-		t.Error("Expected telemetry feature flag to be enabled")
+	if !enabled {
+		t.Error("Expected feature flag to be enabled")
 	}
 }
 
-func TestFetchFeatureFlags_Disabled(t *testing.T) {
+func TestFetchFeatureFlag_Disabled(t *testing.T) {
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 		w.Header().Set("Content-Type", "application/json")
 		w.WriteHeader(http.StatusOK)
-		_, _ = w.Write([]byte(`{"flags": {"databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver": false}}`))
+		_, _ = w.Write([]byte(`{"flags": [{"name": "databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver", "value": "false"}], "ttl_seconds": 300}`))
 	}))
 	defer server.Close()
 
 	host := server.URL // Use full URL for testing
 	httpClient := &http.Client{}
 
-	flags, err := fetchFeatureFlags(context.Background(), host, httpClient)
+	enabled, err := fetchFeatureFlag(context.Background(), host, "test-version", httpClient)
 	if err != nil {
 		t.Errorf("Expected no error, got %v", err)
 	}
-	if flags[flagEnableTelemetry] {
-		t.Error("Expected telemetry feature flag to be disabled")
+	if enabled {
+		t.Error("Expected feature flag to be disabled")
 	}
 }
 
-func TestFetchFeatureFlags_FlagNotPresent(t *testing.T) {
+func TestFetchFeatureFlag_FlagNotPresent(t *testing.T) {
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 		w.Header().Set("Content-Type", "application/json")
 		w.WriteHeader(http.StatusOK)
-		_, _ = w.Write([]byte(`{"flags": {}}`))
+		_, _ = w.Write([]byte(`{"flags": [], "ttl_seconds": 300}`))
 	}))
 	defer server.Close()
 
 	host := server.URL // Use full URL for testing
 	httpClient := &http.Client{}
 
-	flags, err := fetchFeatureFlags(context.Background(), host, httpClient)
+	enabled, err := fetchFeatureFlag(context.Background(), host, "test-version", httpClient)
 	if err != nil {
 		t.Errorf("Expected no error, got %v", err)
 	}
-	if flags[flagEnableTelemetry] {
-		t.Error("Expected telemetry feature flag to be false when not present")
+	if enabled {
+		t.Error("Expected feature flag to be false when not present")
 	}
 }
 
-func TestFetchFeatureFlags_HTTPError(t *testing.T) {
+func TestFetchFeatureFlag_HTTPError(t *testing.T) {
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 		w.WriteHeader(http.StatusInternalServerError)
 	}))
@@ -401,13 +388,13 @@ func TestFetchFeatureFlags_HTTPError(t *testing.T) {
 	host := server.URL // Use full URL for testing
 	httpClient := &http.Client{}
 
-	_, err := fetchFeatureFlags(context.Background(), host, httpClient)
+	_, err := fetchFeatureFlag(context.Background(), host, "test-version", httpClient)
 	if err == nil {
 		t.Error("Expected error for HTTP 500")
 	}
 }
 
-func TestFetchFeatureFlags_InvalidJSON(t *testing.T) {
+func TestFetchFeatureFlag_InvalidJSON(t *testing.T) {
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 		w.Header().Set("Content-Type", "application/json")
 		w.WriteHeader(http.StatusOK)
@@ -418,13 +405,13 @@ func TestFetchFeatureFlags_InvalidJSON(t *testing.T) {
 	host := server.URL // Use full URL for testing
 	httpClient := &http.Client{}
 
-	_, err := fetchFeatureFlags(context.Background(), host, httpClient)
+	_, err := fetchFeatureFlag(context.Background(), host, "test-version", httpClient)
 	if err == nil {
 		t.Error("Expected error for invalid JSON")
 	}
 }
 
-func TestFetchFeatureFlags_ContextCancellation(t *testing.T) {
+func TestFetchFeatureFlag_ContextCancellation(t *testing.T) {
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 		time.Sleep(100 * time.Millisecond)
 		w.WriteHeader(http.StatusOK)
@@ -437,8 +424,13 @@ func TestFetchFeatureFlags_ContextCancellation(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	cancel() // Cancel immediately
 
-	_, err := fetchFeatureFlags(ctx, host, httpClient)
+	_, err := fetchFeatureFlag(ctx, host, "test-version", httpClient)
 	if err == nil {
 		t.Error("Expected error for cancelled context")
 	}
 }
+
+// Helper function to create bool pointer
+func boolPtr(b bool) *bool {
+	return &b
+}
diff --git a/telemetry/integration_test.go b/telemetry/integration_test.go
new file mode 100644
index 00000000..13a6c8b3
--- /dev/null
+++ b/telemetry/integration_test.go
@@ -0,0 +1,311 @@
+package telemetry
+
+import (
+	"context"
+	"encoding/json"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"sync"
+	"sync/atomic"
+	"testing"
+	"time"
+)
+
+// TestIntegration_EndToEnd_WithCircuitBreaker tests complete end-to-end flow.
+func TestIntegration_EndToEnd_WithCircuitBreaker(t *testing.T) {
+	if testing.Short() {
+		t.Skip("skipping integration test in short mode")
+	}
+
+	cfg := DefaultConfig()
+	cfg.FlushInterval = 100 * time.Millisecond
+	cfg.BatchSize = 5
+	httpClient := &http.Client{Timeout: 5 * time.Second}
+
+	requestCount := int32(0)
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		atomic.AddInt32(&requestCount, 1)
+
+		// Verify request structure
+		if r.Method != "POST" {
+			t.Errorf("Expected POST, got %s", r.Method)
+		}
+		if r.URL.Path != "/telemetry-ext" {
+			t.Errorf("Expected /telemetry-ext, got %s", r.URL.Path)
+		}
+
+		// Parse payload (new TelemetryRequest format)
+		body, _ := io.ReadAll(r.Body)
+		var payload TelemetryRequest
+		if err := json.Unmarshal(body, &payload); err != nil {
+			t.Errorf("Failed to parse payload: %v", err)
+		}
+
+		w.WriteHeader(http.StatusOK)
+	}))
+	defer server.Close()
+
+	// Create telemetry client
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
+	aggregator := newMetricsAggregator(exporter, cfg)
+	defer aggregator.close(context.Background()) //nolint:errcheck
+
+	interceptor := newInterceptor(aggregator, true)
+
+	// Simulate statement execution
+	ctx := context.Background()
+	for i := 0; i < 10; i++ {
+		statementID := "stmt-integration"
+		ctx = interceptor.BeforeExecute(ctx, "session-id", statementID)
+		time.Sleep(10 * time.Millisecond) // Simulate work
+		interceptor.AfterExecute(ctx, nil)
+		interceptor.CompleteStatement(ctx, statementID, false)
+	}
+
+	// Wait for flush
+	time.Sleep(200 * time.Millisecond)
+
+	// Verify requests were sent
+	count := atomic.LoadInt32(&requestCount)
+	if count == 0 {
+		t.Error("Expected telemetry requests to be sent")
+	}
+
+	t.Logf("Integration test: sent %d requests", count)
+}
+
+// TestIntegration_CircuitBreakerOpening tests circuit breaker behavior under failures.
+func TestIntegration_CircuitBreakerOpening(t *testing.T) {
+	if testing.Short() {
+		t.Skip("skipping integration test in short mode")
+	}
+
+	cfg := DefaultConfig()
+	cfg.FlushInterval = 50 * time.Millisecond
+	cfg.MaxRetries = 0 // No retries for faster test
+	httpClient := &http.Client{Timeout: 5 * time.Second}
+
+	requestCount := int32(0)
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		atomic.AddInt32(&requestCount, 1)
+		// Always fail to trigger circuit breaker
+		w.WriteHeader(http.StatusInternalServerError)
+	}))
+	defer server.Close()
+
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
+	aggregator := newMetricsAggregator(exporter, cfg)
+	defer aggregator.close(context.Background()) //nolint:errcheck
+
+	interceptor := newInterceptor(aggregator, true)
+	cb := exporter.circuitBreaker
+
+	// Send enough requests to open circuit (need 20+ calls with 50%+ failure rate)
+	ctx := context.Background()
+	for i := 0; i < 50; i++ {
+		statementID := "stmt-circuit"
+		ctx = interceptor.BeforeExecute(ctx, "session-id", statementID)
+		interceptor.AfterExecute(ctx, nil)
+		interceptor.CompleteStatement(ctx, statementID, false)
+
+		// Small delay to ensure each batch is processed
+		time.Sleep(20 * time.Millisecond)
+	}
+
+	// Wait for flush and circuit breaker evaluation
+	time.Sleep(500 * time.Millisecond)
+
+	// Verify circuit opened (may still be closed if not enough failures recorded)
+	state := cb.getState()
+	t.Logf("Circuit breaker state after failures: %v", state)
+
+	// Circuit should eventually open, but timing is async
+	// If not open, at least verify requests were attempted
+	initialCount := atomic.LoadInt32(&requestCount)
+	if initialCount == 0 {
+		t.Error("Expected at least some requests to be sent")
+	}
+
+	// Send more requests - should be dropped if circuit is open
+	for i := 0; i < 10; i++ {
+		statementID := "stmt-dropped"
+		ctx = interceptor.BeforeExecute(ctx, "session-id", statementID)
+		interceptor.AfterExecute(ctx, nil)
+		interceptor.CompleteStatement(ctx, statementID, false)
+	}
+
+	time.Sleep(200 * time.Millisecond)
+
+	finalCount := atomic.LoadInt32(&requestCount)
+	t.Logf("Circuit breaker test: %d requests sent, state=%v", finalCount, cb.getState())
+
+	// Test passes if either:
+	// 1. Circuit opened and requests were dropped, OR
+	// 2. Circuit is still trying (which is also acceptable for async system)
+	if state == stateOpen && finalCount > initialCount+5 {
+		t.Errorf("Expected requests to be dropped when circuit open, got %d additional requests", finalCount-initialCount)
+	}
+}
+
+// TestIntegration_OptInPriority_ExplicitOptOut tests explicit opt-out.
+func TestIntegration_OptInPriority_ExplicitOptOut(t *testing.T) {
+	cfg := &Config{
+		EnableTelemetry: false, // Priority 1 (client): Explicit opt-out
+		BatchSize:       100,
+		FlushInterval:   5 * time.Second,
+		MaxRetries:      3,
+		RetryDelay:      100 * time.Millisecond,
+	}
+
+	httpClient := &http.Client{Timeout: 5 * time.Second}
+
+	// Server that returns enabled
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		resp := map[string]interface{}{
+			"flags": map[string]bool{
+				"databricks.partnerplatform.clientConfigsFeatureFlags.enableTelemetryForGoDriver": true,
+			},
+		}
+		_ = json.NewEncoder(w).Encode(resp)
+	}))
+	defer server.Close()
+
+	ctx := context.Background()
+
+	// Should be disabled due to explicit opt-out
+	result := isTelemetryEnabled(ctx, cfg, server.URL, "test-version", httpClient)
+
+	if result {
+		t.Error("Expected telemetry to be disabled by explicit opt-out")
+	}
+}
+
+// TestIntegration_PrivacyCompliance verifies no sensitive data is collected.
+func TestIntegration_PrivacyCompliance_NoQueryText(t *testing.T) {
+	cfg := DefaultConfig()
+	cfg.FlushInterval = 50 * time.Millisecond
+	httpClient := &http.Client{Timeout: 5 * time.Second}
+
+	var mu sync.Mutex
+	var capturedBody []byte
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		body, _ := io.ReadAll(r.Body)
+		mu.Lock()
+		capturedBody = body
+		mu.Unlock()
+		w.WriteHeader(http.StatusOK)
+	}))
+	defer server.Close()
+
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
+	aggregator := newMetricsAggregator(exporter, cfg)
+	defer aggregator.close(context.Background()) //nolint:errcheck
+
+	interceptor := newInterceptor(aggregator, true)
+
+	ctx := context.Background()
+	statementID := "stmt-privacy"
+	ctx = interceptor.BeforeExecute(ctx, "session-id", statementID)
+
+	// Add sensitive tags — none of these should appear in the exported telemetry
+	interceptor.AddTag(ctx, "query.text", "SELECT * FROM users")
+	interceptor.AddTag(ctx, "user.email", "user@example.com")
+
+	interceptor.AfterExecute(ctx, nil)
+	interceptor.CompleteStatement(ctx, statementID, false)
+
+	// Wait for flush
+	time.Sleep(200 * time.Millisecond)
+
+	mu.Lock()
+	body := capturedBody
+	mu.Unlock()
+
+	if len(body) == 0 {
+		t.Fatal("Expected telemetry request to be sent")
+	}
+
+	// The exporter sends TelemetryRequest with ProtoLogs (JSON-encoded TelemetryFrontendLog).
+	// Verify sensitive values are absent from the serialised payload.
+	bodyStr := string(body)
+	if strings.Contains(bodyStr, "SELECT * FROM users") {
+		t.Error("Query text must not be exported")
+	}
+	if strings.Contains(bodyStr, "user@example.com") {
+		t.Error("User email must not be exported")
+	}
+
+	t.Log("Privacy compliance test passed: sensitive data not present in payload")
+}
+
+// TestIntegration_FieldMapping verifies that only known metric fields are exported
+// in the TelemetryRequest format (no generic tag pass-through).
+func TestIntegration_FieldMapping(t *testing.T) {
+	cfg := DefaultConfig()
+	cfg.FlushInterval = 50 * time.Millisecond
+	httpClient := &http.Client{Timeout: 5 * time.Second}
+
+	var capturedRequest TelemetryRequest
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		body, _ := io.ReadAll(r.Body)
+		_ = json.Unmarshal(body, &capturedRequest)
+		w.WriteHeader(http.StatusOK)
+	}))
+	defer server.Close()
+
+	exporter := newTelemetryExporter(server.URL, "test-version", httpClient, cfg)
+
+	metric := &telemetryMetric{
+		metricType:  "connection",
+		timestamp:   time.Now(),
+		workspaceID: "ws-test",
+		sessionID:   "sess-1",
+		latencyMs:   42,
+		tags: map[string]interface{}{
+			"chunk_count":      3,
+			"bytes_downloaded": int64(1024),
+			"unknown.tag":      "value", // should NOT appear in output
+		},
+	}
+
+	ctx := context.Background()
+	exporter.export(ctx, []*telemetryMetric{metric})
+
+	time.Sleep(150 * time.Millisecond)
+
+	if len(capturedRequest.ProtoLogs) == 0 {
+		t.Fatal("Expected at least one ProtoLog entry")
+	}
+
+	// Each ProtoLog entry is a JSON-encoded TelemetryFrontendLog.
+	var log TelemetryFrontendLog
+	if err := json.Unmarshal([]byte(capturedRequest.ProtoLogs[0]), &log); err != nil {
+		t.Fatalf("Failed to unmarshal ProtoLog: %v", err)
+	}
+
+	if log.Entry == nil || log.Entry.SQLDriverLog == nil {
+		t.Fatal("Expected SQLDriverLog to be populated")
+	}
+
+	entry := log.Entry.SQLDriverLog
+	if entry.SessionID != "sess-1" {
+		t.Errorf("Expected session_id=sess-1, got %q", entry.SessionID)
+	}
+	if entry.OperationLatencyMs != 42 {
+		t.Errorf("Expected latency=42, got %d", entry.OperationLatencyMs)
+	}
+	if entry.SQLOperation != nil && entry.SQLOperation.ChunkDetails != nil {
+		if entry.SQLOperation.ChunkDetails.TotalChunksIterated != 3 {
+			t.Errorf("Expected total_chunks_iterated=3, got %d", entry.SQLOperation.ChunkDetails.TotalChunksIterated)
+		}
+	}
+
+	// unknown.tag must not appear anywhere in the serialised output
+	if strings.Contains(capturedRequest.ProtoLogs[0], "unknown.tag") {
+		t.Error("unknown.tag must not be exported")
+	}
+
+	t.Log("Field mapping test passed")
+}
diff --git a/telemetry/interceptor.go b/telemetry/interceptor.go
index 419b4900..5ef01b1a 100644
--- a/telemetry/interceptor.go
+++ b/telemetry/interceptor.go
@@ -166,6 +166,34 @@ func (i *Interceptor) CompleteStatement(ctx context.Context, statementID string,
 	i.aggregator.completeStatement(ctx, statementID, failed)
 }
 
+// RecordOperation records an operation with type, latency, and optional error.
+// Exported for use by the driver package.
+func (i *Interceptor) RecordOperation(ctx context.Context, sessionID string, operationType string, latencyMs int64, err error) {
+	if !i.enabled {
+		return
+	}
+
+	defer func() {
+		if r := recover(); r != nil {
+			logger.Trace().Msgf("telemetry: recordOperation panic: %v", r)
+		}
+	}()
+
+	metric := &telemetryMetric{
+		metricType: "operation",
+		timestamp:  time.Now(),
+		sessionID:  sessionID,
+		latencyMs:  latencyMs,
+		tags:       map[string]interface{}{"operation_type": operationType},
+	}
+
+	if err != nil {
+		metric.errorType = classifyError(err)
+	}
+
+	i.aggregator.recordMetric(ctx, metric)
+}
+
 // Close flushes any pending per-connection metrics.
 // Does NOT close the shared aggregator — its lifecycle is managed via
 // ReleaseForConnection, which uses reference counting across all connections
@@ -176,6 +204,12 @@ func (i *Interceptor) Close(ctx context.Context) error {
 		return nil
 	}
 
-	i.aggregator.flush(ctx)
+	defer func() {
+		if r := recover(); r != nil {
+			logger.Debug().Msgf("telemetry: Close panic: %v", r)
+		}
+	}()
+
+	i.aggregator.flushSync(ctx)
 	return nil
 }
diff --git a/telemetry/manager.go b/telemetry/manager.go
index 33bfe1cf..8977e924 100644
--- a/telemetry/manager.go
+++ b/telemetry/manager.go
@@ -45,13 +45,13 @@ func getClientManager() *clientManager {
 
 // getOrCreateClient gets or creates a telemetry client for the host.
 // Increments reference count.
-func (m *clientManager) getOrCreateClient(host string, httpClient *http.Client, cfg *Config) *telemetryClient {
+func (m *clientManager) getOrCreateClient(host string, driverVersion string, httpClient *http.Client, cfg *Config) *telemetryClient {
 	m.mu.Lock()
 	defer m.mu.Unlock()
 
 	holder, exists := m.clients[host]
 	if !exists {
-		client := newTelemetryClient(host, httpClient, cfg)
+		client := newTelemetryClient(host, driverVersion, httpClient, cfg)
 		if err := client.start(); err != nil {
 			// Failed to start client, don't add to map
 			logger.Logger.Debug().Str("host", host).Err(err).Msg("failed to start telemetry client")
diff --git a/telemetry/manager_test.go b/telemetry/manager_test.go
index 59127e24..51461452 100644
--- a/telemetry/manager_test.go
+++ b/telemetry/manager_test.go
@@ -29,7 +29,7 @@ func TestClientManager_GetOrCreateClient(t *testing.T) {
 	cfg := DefaultConfig()
 
 	// First call should create client and increment refCount to 1
-	client1 := manager.getOrCreateClient(host, httpClient, cfg)
+	client1 := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 	if client1 == nil {
 		t.Fatal("Expected client to be created")
 	}
@@ -46,7 +46,7 @@ func TestClientManager_GetOrCreateClient(t *testing.T) {
 	}
 
 	// Second call should reuse client and increment refCount to 2
-	client2 := manager.getOrCreateClient(host, httpClient, cfg)
+	client2 := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 	if client2 != client1 {
 		t.Error("Expected to get the same client instance")
 	}
@@ -65,8 +65,8 @@ func TestClientManager_GetOrCreateClient_DifferentHosts(t *testing.T) {
 	httpClient := &http.Client{}
 	cfg := DefaultConfig()
 
-	client1 := manager.getOrCreateClient(host1, httpClient, cfg)
-	client2 := manager.getOrCreateClient(host2, httpClient, cfg)
+	client1 := manager.getOrCreateClient(host1, "test-version", httpClient, cfg)
+	client2 := manager.getOrCreateClient(host2, "test-version", httpClient, cfg)
 
 	if client1 == client2 {
 		t.Error("Expected different clients for different hosts")
@@ -87,8 +87,8 @@ func TestClientManager_ReleaseClient(t *testing.T) {
 	cfg := DefaultConfig()
 
 	// Create client with refCount = 2
-	manager.getOrCreateClient(host, httpClient, cfg)
-	manager.getOrCreateClient(host, httpClient, cfg)
+	manager.getOrCreateClient(host, "test-version", httpClient, cfg)
+	manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 
 	// First release should decrement to 1
 	err := manager.releaseClient(host)
@@ -151,7 +151,7 @@ func TestClientManager_ConcurrentAccess(t *testing.T) {
 	for i := 0; i < numGoroutines; i++ {
 		go func() {
 			defer wg.Done()
-			client := manager.getOrCreateClient(host, httpClient, cfg)
+			client := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 			if client == nil {
 				t.Error("Expected client to be created")
 			}
@@ -207,7 +207,7 @@ func TestClientManager_ConcurrentAccessMultipleHosts(t *testing.T) {
 			wg.Add(1)
 			go func(h string) {
 				defer wg.Done()
-				_ = manager.getOrCreateClient(h, httpClient, cfg)
+				_ = manager.getOrCreateClient(h, "test-version", httpClient, cfg)
 			}(host)
 		}
 	}
@@ -241,7 +241,7 @@ func TestClientManager_ReleaseClientPartial(t *testing.T) {
 
 	// Create 5 references
 	for i := 0; i < 5; i++ {
-		manager.getOrCreateClient(host, httpClient, cfg)
+		manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 	}
 
 	// Release 3 references
@@ -271,7 +271,7 @@ func TestClientManager_ClientStartCalled(t *testing.T) {
 	httpClient := &http.Client{}
 	cfg := DefaultConfig()
 
-	client := manager.getOrCreateClient(host, httpClient, cfg)
+	client := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 
 	if !client.started {
 		t.Error("Expected start() to be called on new client")
@@ -287,7 +287,7 @@ func TestClientManager_ClientCloseCalled(t *testing.T) {
 	httpClient := &http.Client{}
 	cfg := DefaultConfig()
 
-	client := manager.getOrCreateClient(host, httpClient, cfg)
+	client := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 	_ = manager.releaseClient(host)
 
 	if !client.closed {
@@ -305,9 +305,9 @@ func TestClientManager_MultipleGetOrCreateSameClient(t *testing.T) {
 	cfg := DefaultConfig()
 
 	// Get same client multiple times
-	client1 := manager.getOrCreateClient(host, httpClient, cfg)
-	client2 := manager.getOrCreateClient(host, httpClient, cfg)
-	client3 := manager.getOrCreateClient(host, httpClient, cfg)
+	client1 := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
+	client2 := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
+	client3 := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 
 	// All should be same instance
 	if client1 != client2 || client2 != client3 {
@@ -337,7 +337,7 @@ func TestClientManager_Shutdown(t *testing.T) {
 	// Create clients for multiple hosts
 	clients := make([]*telemetryClient, 0, len(hosts))
 	for _, host := range hosts {
-		client := manager.getOrCreateClient(host, httpClient, cfg)
+		client := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 		clients = append(clients, client)
 	}
 
@@ -375,9 +375,9 @@ func TestClientManager_ShutdownWithActiveRefs(t *testing.T) {
 	cfg := DefaultConfig()
 
 	// Create client with multiple references
-	client := manager.getOrCreateClient(host, httpClient, cfg)
-	manager.getOrCreateClient(host, httpClient, cfg)
-	manager.getOrCreateClient(host, httpClient, cfg)
+	client := manager.getOrCreateClient(host, "test-version", httpClient, cfg)
+	manager.getOrCreateClient(host, "test-version", httpClient, cfg)
+	manager.getOrCreateClient(host, "test-version", httpClient, cfg)
 
 	holder := manager.clients[host]
 	if holder.refCount != 3 {
diff --git a/telemetry/operation_type.go b/telemetry/operation_type.go
new file mode 100644
index 00000000..3ba6f213
--- /dev/null
+++ b/telemetry/operation_type.go
@@ -0,0 +1,16 @@
+package telemetry
+
+const (
+	OperationTypeUnspecified      = "TYPE_UNSPECIFIED"
+	OperationTypeCreateSession    = "CREATE_SESSION"
+	OperationTypeDeleteSession    = "DELETE_SESSION"
+	OperationTypeExecuteStatement = "EXECUTE_STATEMENT"
+	OperationTypeCloseStatement   = "CLOSE_STATEMENT"
+)
+
+// isTerminalOperationType returns true for operations that signal the end of a
+// session or statement lifecycle. These flush immediately so metrics are not
+// lost if the connection closes before the next periodic flush — matching JDBC.
+func isTerminalOperationType(opType string) bool {
+	return opType == OperationTypeDeleteSession || opType == OperationTypeCloseStatement
+}
diff --git a/telemetry/request.go b/telemetry/request.go
new file mode 100644
index 00000000..f516406d
--- /dev/null
+++ b/telemetry/request.go
@@ -0,0 +1,228 @@
+package telemetry
+
+import (
+	"crypto/rand"
+	"encoding/hex"
+	"encoding/json"
+	"time"
+)
+
+// TelemetryRequest is the top-level request sent to the telemetry endpoint.
+type TelemetryRequest struct {
+	UploadTime int64    `json:"uploadTime"`
+	Items      []string `json:"items"`
+	ProtoLogs  []string `json:"protoLogs"`
+}
+
+// TelemetryFrontendLog represents a single telemetry log entry.
+type TelemetryFrontendLog struct {
+	WorkspaceID        int64               `json:"workspace_id,omitempty"`
+	FrontendLogEventID string              `json:"frontend_log_event_id,omitempty"`
+	Context            *FrontendLogContext `json:"context,omitempty"`
+	Entry              *FrontendLogEntry   `json:"entry,omitempty"`
+}
+
+// FrontendLogContext contains the client context.
+type FrontendLogContext struct {
+	ClientContext *TelemetryClientContext `json:"client_context,omitempty"`
+}
+
+// TelemetryClientContext contains client-level information.
+type TelemetryClientContext struct {
+	ClientType    string `json:"client_type,omitempty"`
+	ClientVersion string `json:"client_version,omitempty"`
+}
+
+// FrontendLogEntry contains the actual telemetry event.
+type FrontendLogEntry struct {
+	SQLDriverLog *TelemetryEvent `json:"sql_driver_log,omitempty"`
+}
+
+// TelemetryEvent maps to OssSqlDriverTelemetryLog in the proto schema.
+type TelemetryEvent struct {
+	SessionID                  string                      `json:"session_id,omitempty"`
+	SQLStatementID             string                      `json:"sql_statement_id,omitempty"`
+	SystemConfiguration        *DriverSystemConfiguration  `json:"system_configuration,omitempty"`
+	DriverConnectionParameters *DriverConnectionParameters `json:"driver_connection_params,omitempty"`
+	AuthType                   string                      `json:"auth_type,omitempty"`
+	VolumeOperation            *VolumeOperationEvent       `json:"vol_operation,omitempty"`
+	SQLOperation               *SQLExecutionEvent          `json:"sql_operation,omitempty"`
+	ErrorInfo                  *DriverErrorInfo            `json:"error_info,omitempty"`
+	OperationLatencyMs         int64                       `json:"operation_latency_ms,omitempty"`
+}
+
+// DriverSystemConfiguration maps to DriverSystemConfiguration in the proto schema.
+type DriverSystemConfiguration struct {
+	DriverVersion   string `json:"driver_version,omitempty"`
+	RuntimeName     string `json:"runtime_name,omitempty"`
+	RuntimeVersion  string `json:"runtime_version,omitempty"`
+	RuntimeVendor   string `json:"runtime_vendor,omitempty"`
+	OSName          string `json:"os_name,omitempty"`
+	OSVersion       string `json:"os_version,omitempty"`
+	OSArch          string `json:"os_arch,omitempty"`
+	DriverName      string `json:"driver_name,omitempty"`
+	ClientAppName   string `json:"client_app_name,omitempty"`
+	LocaleName      string `json:"locale_name,omitempty"`
+	CharSetEncoding string `json:"char_set_encoding,omitempty"`
+	ProcessName     string `json:"process_name,omitempty"`
+}
+
+// HostDetails maps to HostDetails in the proto schema.
+type HostDetails struct {
+	HostURL       string `json:"host_url,omitempty"`
+	Port          int32  `json:"port,omitempty"`
+	ProxyAuthType string `json:"proxy_auth_type,omitempty"`
+}
+
+// DriverConnectionParameters maps to DriverConnectionParameters in the proto schema.
+// Only fields populated by the Go driver are included; others are omitted.
+type DriverConnectionParameters struct {
+	HTTPPath             string       `json:"http_path,omitempty"`
+	Mode                 string       `json:"mode,omitempty"`
+	HostInfo             *HostDetails `json:"host_info,omitempty"`
+	UseProxy             bool         `json:"use_proxy,omitempty"`
+	AuthMech             string       `json:"auth_mech,omitempty"`
+	AuthFlow             string       `json:"auth_flow,omitempty"`
+	AuthScope            string       `json:"auth_scope,omitempty"`
+	UseSystemProxy       bool         `json:"use_system_proxy,omitempty"`
+	UseCFProxy           bool         `json:"use_cf_proxy,omitempty"`
+	EnableArrow          bool         `json:"enable_arrow,omitempty"`
+	EnableDirectResults  bool         `json:"enable_direct_results,omitempty"`
+	QueryTags            string       `json:"query_tags,omitempty"`
+	EnableMetricViewMeta bool         `json:"enable_metric_view_metadata,omitempty"`
+	SocketTimeout        int64        `json:"socket_timeout,omitempty"`
+}
+
+// SQLExecutionEvent maps to SqlExecutionEvent in the proto schema.
+type SQLExecutionEvent struct {
+	StatementType        string           `json:"statement_type,omitempty"`
+	IsCompressed         bool             `json:"is_compressed,omitempty"`
+	ExecutionResult      string           `json:"execution_result,omitempty"`
+	ChunkID              int64            `json:"chunk_id,omitempty"`
+	RetryCount           int64            `json:"retry_count,omitempty"`
+	ChunkDetails         *ChunkDetails    `json:"chunk_details,omitempty"`
+	ResultLatency        *ResultLatency   `json:"result_latency,omitempty"`
+	OperationDetail      *OperationDetail `json:"operation_detail,omitempty"`
+	JavaUsesPatchedArrow bool             `json:"java_uses_patched_arrow,omitempty"`
+}
+
+// ChunkDetails maps to ChunkDetails in the proto schema.
+type ChunkDetails struct {
+	InitialChunkLatencyMs   int64 `json:"initial_chunk_latency_millis,omitempty"`
+	SlowestChunkLatencyMs   int64 `json:"slowest_chunk_latency_millis,omitempty"`
+	TotalChunksPresent      int32 `json:"total_chunks_present,omitempty"`
+	TotalChunksIterated     int32 `json:"total_chunks_iterated,omitempty"`
+	SumChunksDownloadTimeMs int64 `json:"sum_chunks_download_time_millis,omitempty"`
+}
+
+// ResultLatency maps to ResultLatency in the proto schema.
+type ResultLatency struct {
+	ResultSetReadyLatencyMs       int64 `json:"result_set_ready_latency_millis,omitempty"`
+	ResultSetConsumptionLatencyMs int64 `json:"result_set_consumption_latency_millis,omitempty"`
+}
+
+// OperationDetail maps to OperationDetail in the proto schema.
+type OperationDetail struct {
+	NOperationStatusCalls    int32  `json:"n_operation_status_calls,omitempty"`
+	OperationStatusLatencyMs int64  `json:"operation_status_latency_millis,omitempty"`
+	OperationType            string `json:"operation_type,omitempty"`
+	IsInternalCall           bool   `json:"is_internal_call,omitempty"`
+}
+
+// VolumeOperationEvent maps to VolumeOperationEvent in the proto schema.
+type VolumeOperationEvent struct {
+	VolumeOperationType string `json:"volume_operation_type,omitempty"`
+	VolumePath          string `json:"volume_path,omitempty"`
+	LocalFile           string `json:"local_file,omitempty"`
+}
+
+// DriverErrorInfo maps to DriverErrorInfo in the proto schema.
+type DriverErrorInfo struct {
+	ErrorName  string `json:"error_name,omitempty"`
+	StackTrace string `json:"stack_trace,omitempty"`
+}
+
+// TelemetryResponse is the response from the telemetry endpoint.
+type TelemetryResponse struct {
+	Errors             []string `json:"errors"`
+	NumSuccess         int      `json:"numSuccess"`
+	NumProtoSuccess    int      `json:"numProtoSuccess"`
+	NumRealtimeSuccess int      `json:"numRealtimeSuccess"`
+}
+
+// createTelemetryRequest creates a telemetry request from metrics.
+func createTelemetryRequest(metrics []*telemetryMetric, driverVersion string) (*TelemetryRequest, error) {
+	protoLogs := make([]string, 0, len(metrics))
+
+	for _, metric := range metrics {
+		frontendLog := &TelemetryFrontendLog{
+			FrontendLogEventID: generateEventID(),
+			Context: &FrontendLogContext{
+				ClientContext: &TelemetryClientContext{
+					ClientType:    "golang",
+					ClientVersion: driverVersion,
+				},
+			},
+			Entry: &FrontendLogEntry{
+				SQLDriverLog: &TelemetryEvent{
+					SessionID:           metric.sessionID,
+					SQLStatementID:      metric.statementID,
+					SystemConfiguration: getSystemConfiguration(driverVersion),
+					OperationLatencyMs:  metric.latencyMs,
+				},
+			},
+		}
+
+		// Add SQL operation details if available.
+		if tags := metric.tags; tags != nil {
+			sqlOp := &SQLExecutionEvent{}
+
+			if v, ok := tags["result.format"].(string); ok {
+				sqlOp.ExecutionResult = v
+			}
+			if chunkCount, ok := tags["chunk_count"].(int); ok && chunkCount > 0 {
+				sqlOp.ChunkDetails = &ChunkDetails{
+					TotalChunksIterated: int32(chunkCount), //nolint:gosec // chunk count is always small
+				}
+			}
+
+			if opType, ok := tags["operation_type"].(string); ok {
+				detail := &OperationDetail{
+					OperationType: opType,
+				}
+				if pollCount, ok := tags["poll_count"].(int); ok {
+					detail.NOperationStatusCalls = int32(pollCount) //nolint:gosec // poll count is always small
+				}
+				sqlOp.OperationDetail = detail
+			}
+
+			frontendLog.Entry.SQLDriverLog.SQLOperation = sqlOp
+		}
+
+		// Add error info if present.
+		if metric.errorType != "" {
+			frontendLog.Entry.SQLDriverLog.ErrorInfo = &DriverErrorInfo{
+				ErrorName: metric.errorType,
+			}
+		}
+
+		jsonBytes, err := json.Marshal(frontendLog)
+		if err != nil {
+			return nil, err
+		}
+		protoLogs = append(protoLogs, string(jsonBytes))
+	}
+
+	return &TelemetryRequest{
+		UploadTime: time.Now().UnixMilli(),
+		Items:      []string{}, // Required but empty
+		ProtoLogs:  protoLogs,
+	}, nil
+}
+
+// generateEventID generates a unique event ID using crypto/rand.
+func generateEventID() string {
+	b := make([]byte, 8)
+	_, _ = rand.Read(b)
+	return time.Now().Format("20060102150405") + "-" + hex.EncodeToString(b)
+}
diff --git a/telemetry/system_info.go b/telemetry/system_info.go
new file mode 100644
index 00000000..56b979b9
--- /dev/null
+++ b/telemetry/system_info.go
@@ -0,0 +1,110 @@
+package telemetry
+
+import (
+	"os"
+	"runtime"
+	"strings"
+	"sync"
+)
+
+// sysInfoOnce caches the parts of system configuration that are invariant across calls
+// (OS info, runtime, process name) to avoid repeated os.ReadFile on every metric.
+var (
+	sysInfoOnce   sync.Once
+	cachedOSName  string
+	cachedOSVer   string
+	cachedArch    string
+	cachedRuntime string
+	cachedLocale  string
+	cachedProcess string
+)
+
+func initSysInfo() {
+	sysInfoOnce.Do(func() {
+		cachedOSName = getOSName()
+		cachedOSVer = getOSVersion()
+		cachedArch = runtime.GOARCH
+		cachedRuntime = runtime.Version()
+		cachedLocale = getLocaleName()
+		cachedProcess = getProcessName()
+	})
+}
+
+func getSystemConfiguration(driverVersion string) *DriverSystemConfiguration {
+	initSysInfo()
+	return &DriverSystemConfiguration{
+		OSName:          cachedOSName,
+		OSVersion:       cachedOSVer,
+		OSArch:          cachedArch,
+		DriverName:      "databricks-sql-go",
+		DriverVersion:   driverVersion,
+		RuntimeName:     "go",
+		RuntimeVersion:  cachedRuntime,
+		RuntimeVendor:   "",
+		LocaleName:      cachedLocale,
+		CharSetEncoding: "UTF-8",
+		ProcessName:     cachedProcess,
+	}
+}
+
+func getOSName() string {
+	switch runtime.GOOS {
+	case "darwin":
+		return "macOS"
+	case "windows":
+		return "Windows"
+	case "linux":
+		return "Linux"
+	default:
+		return runtime.GOOS
+	}
+}
+
+func getOSVersion() string {
+	switch runtime.GOOS {
+	case "linux":
+		if data, err := os.ReadFile("/etc/os-release"); err == nil {
+			lines := strings.Split(string(data), "\n")
+			for _, line := range lines {
+				if strings.HasPrefix(line, "VERSION=") {
+					version := strings.TrimPrefix(line, "VERSION=")
+					version = strings.Trim(version, "\"")
+					return version
+				}
+			}
+		}
+		if data, err := os.ReadFile("/proc/version"); err == nil {
+			return strings.Split(string(data), " ")[2]
+		}
+	}
+	return ""
+}
+
+func getLocaleName() string {
+	if lang := os.Getenv("LANG"); lang != "" {
+		parts := strings.Split(lang, ".")
+		if len(parts) > 0 {
+			return parts[0]
+		}
+	}
+	return "en_US"
+}
+
+func getProcessName() string {
+	if len(os.Args) > 0 {
+		processPath := os.Args[0]
+		lastSlash := strings.LastIndex(processPath, "/")
+		if lastSlash == -1 {
+			lastSlash = strings.LastIndex(processPath, "\\")
+		}
+		if lastSlash >= 0 {
+			processPath = processPath[lastSlash+1:]
+		}
+		dotIndex := strings.LastIndex(processPath, ".")
+		if dotIndex > 0 {
+			processPath = processPath[:dotIndex]
+		}
+		return processPath
+	}
+	return ""
+}