obs(grafana): add Pebble internals + unsupported-commands panel; refactor Redis/Cluster dashboards

README.md

@@ -52,16 +52,20 @@ Provisioned monitoring assets live under:
 
 - `monitoring/prometheus/prometheus.yml`
 - `monitoring/grafana/dashboards/elastickv-cluster-overview.json`
-- `monitoring/grafana/dashboards/elastickv-cluster-summary.json`
+- `monitoring/grafana/dashboards/elastickv-dynamodb.json`
 - `monitoring/grafana/dashboards/elastickv-raft-status.json`
+- `monitoring/grafana/dashboards/elastickv-redis-summary.json`
+- `monitoring/grafana/dashboards/elastickv-pebble-internals.json`
 - `monitoring/grafana/provisioning/`
 - `monitoring/docker-compose.yml`
 
 The provisioned dashboards are organized by operator task:
 
-- `Elastickv Cluster Overview` is the landing page for leader identity, cluster-wide latency/error posture, and per-node Raft health
-- `Elastickv Request Health` is the DynamoDB/API drilldown for slow operations, noisy nodes, and hot/erroring tables
+- `Elastickv Cluster` is the landing page for leader identity, cluster-wide latency/error posture, and per-node Raft health
+- `Elastickv DynamoDB` is the DynamoDB-compatible API drilldown for slow operations, noisy nodes, and hot/erroring tables
 - `Elastickv Raft Status` is the control-plane drilldown for membership, leader changes, failed proposals, node state, index drift, backlog, and leader contact
+- `Elastickv Redis` is the Redis-compatible API drilldown for per-command throughput/latency/errors, with a collapsible `Hot Path` row for GET fast-path (PR #560) verification
+- `Elastickv Pebble Internals` is the storage-engine drilldown for block cache, L0 pressure, compactions, memtables, and store write conflicts
 
 If you bind `--metricsAddress` to a non-loopback address, `--metricsToken` is required. Prometheus must send the same bearer token, for example:
 

docs/redis_hotpath_dashboard.md

@@ -1,9 +1,11 @@
 # Redis Hot Path Dashboard (PR #560 verification)
 
-`monitoring/grafana/dashboards/elastickv-redis-hotpath.json` is the
+The "Hot Path (legacy PR #560)" collapsed row at the bottom of
+`monitoring/grafana/dashboards/elastickv-redis-summary.json` is the
 operator view for the Redis GET hot path. It was added to confirm that
 PR #560 (`a45ca291` "perf(redis): fast-path GET to avoid ~17-seek type
-probe") landed cleanly in production.
+probe") landed cleanly in production. Expand the row to see the
+panels described below.
 
 ## How to confirm #560 worked
 

internal/raftengine/etcd/engine.go

@@ -750,7 +750,7 @@ func (e *Engine) LastQuorumAck() time.Time {
 // heartbeat channel was full. Monotonic across the life of the engine.
 // Surfaced to Prometheus via the monitoring package so the hot-path
 // dashboard can graph stepCh saturation alongside LinearizableRead
-// rate (see monitoring/grafana/dashboards/elastickv-redis-hotpath.json).
+// rate (see the "Hot Path" row in monitoring/grafana/dashboards/elastickv-redis-summary.json).
 func (e *Engine) DispatchDropCount() uint64 {
 	if e == nil {
 		return 0

kv/coordinator.go

@@ -60,7 +60,7 @@ type LeaseReadObserver interface {
 // WithLeaseReadObserver wires a LeaseReadObserver onto a Coordinate.
 // This is the mechanism monitoring uses to surface the lease-hit ratio
 // panel on the Redis hot-path dashboard (see
-// monitoring/grafana/dashboards/elastickv-redis-hotpath.json).
+// the "Hot Path" row in monitoring/grafana/dashboards/elastickv-redis-summary.json).
 //
 // Typed-nil guard: a caller passing a typed-nil pointer
 // (e.g. `var o *myObserver; WithLeaseReadObserver(o)`) produces an

monitoring/grafana/dashboards/elastickv-cluster-overview.json

@@ -1280,7 +1280,7 @@
   },
   "timepicker": {},
   "timezone": "browser",
-  "title": "Elastickv Cluster Overview",
+  "title": "Elastickv Cluster",
   "uid": "elastickv-cluster",
   "version": 2
 }

monitoring/grafana/dashboards/elastickv-cluster-summary.json → monitoring/grafana/dashboards/elastickv-dynamodb.json

@@ -15,7 +15,7 @@
       }
     ]
   },
-  "description": "DynamoDB-compatible request health for elastickv: throughput, latency, errors, tables, and noisy nodes.",
+  "description": "DynamoDB-compatible API health for elastickv: per-operation throughput and latency, per-table request/error breakdown, and item-volume panels.",
   "editable": true,
   "fiscalYearStartMonth": 0,
   "graphTooltip": 1,
@@ -918,7 +918,7 @@
     "elastickv",
     "dynamodb",
     "requests",
-    "summary"
+    "api"
   ],
   "templating": {
     "list": [
@@ -1019,7 +1019,7 @@
   },
   "timepicker": {},
   "timezone": "browser",
-  "title": "Elastickv Request Health",
-  "uid": "elastickv-cluster-summary",
+  "title": "Elastickv DynamoDB",
+  "uid": "elastickv-dynamodb",
   "version": 2
 }