Add data lake best practices guide#6512
Open
amychen1776 wants to merge 6 commits into
Open
Conversation
Correct YAML frontmatter, use event_time in the virtual-column debug example for Iceberg hidden partitioning, and update the data lake index blurb. Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
3 Skipped Deployments
|
Remove nested backticks that break inline code parsing and show SHOW CREATE TABLE in a proper code block. Co-authored-by: Cursor <cursoragent@cursor.com>
dhtclk
reviewed
Jul 8, 2026
Collapse consecutive blank lines around tables and headings (MD012) and add "Delta Kernel" to the Headings.yml sentence-case exceptions so the heading keeps the proper project name. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
PREWHERE -> /optimize/prewhere and filesystem cache disk -> /operations/storing-data#using-local-cache. These surfaced once the markdown-lint gate passed and the Docusaurus build could run. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
dhtclk
reviewed
Jul 9, 2026
|
|
||
| Version numbers on this page match ClickHouse release versions (Cloud and self-managed). Check your service version before enabling a setting or feature. | ||
|
|
||
| Lake query performance depends on how much metadata and how many [Parquet](/interfaces/formats/Parquet) files ClickHouse reads from object storage. As with ClickHouse, query performance is improved by filtering on partition columns and selecting fewer columns. |
Collaborator
There was a problem hiding this comment.
Suggested change
| Lake query performance depends on how much metadata and how many [Parquet](/interfaces/formats/Parquet) files ClickHouse reads from object storage. As with ClickHouse, query performance is improved by filtering on partition columns and selecting fewer columns. | |
| Lake query performance depends on how much metadata and how many [Parquet](/interfaces/formats/Parquet) files ClickHouse reads from object storage. As with any ClickHouse table, query performance is improved by filtering on partition columns and selecting fewer columns. |
Slightly odd phrasing here for a ClickHouse doc. Tried rewording slightly so it doesn't read like ClickHouse is a separate thing.
dhtclk
reviewed
Jul 9, 2026
|
|
||
| ### Parallel reads on multi-node clusters {#parallel-cluster-reads} | ||
|
|
||
| On ClickHouse Cloud and self-managed multi-node services, cluster variants of lake table functions distribute [Parquet](/interfaces/formats/Parquet) file reads across replicas. The initiator node dispatches files to workers in parallel. Use cluster variants for batch reads and scheduled loads over large tables. On single-node deployments, the standard table function is enough. |
Collaborator
There was a problem hiding this comment.
Suggested change
| On ClickHouse Cloud and self-managed multi-node services, cluster variants of lake table functions distribute [Parquet](/interfaces/formats/Parquet) file reads across replicas. The initiator node dispatches files to workers in parallel. Use cluster variants for batch reads and scheduled loads over large tables. On single-node deployments, the standard table function is enough. | |
| On ClickHouse Cloud and self-managed multi-node services, cluster variants of lake table functions distribute [Parquet](/interfaces/formats/Parquet) file reads across replicas. The initiator node dispatches files to workers in parallel. Use cluster variants for batch reads and scheduled loads over large tables. On single-node deployments, the standard table function is enough. |
dhtclk
reviewed
Jul 9, 2026
|
|
||
| You can combine cluster reads with other performance settings. | ||
|
|
||
| ### Bound batch reads to snapshots {#snapshot-bounds} |
Collaborator
There was a problem hiding this comment.
This is non-blocking, but I find Bound batch reads to snapshots a bit awkward wording for a header on first read. Perhaps "Scope each run to snapshot range"?
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
/use-cases/data-lake/best-practicescovering access methods (table functions, table engines,DataLakeCatalog), required feature flags, and performance tuning for Iceberg and Delta Lake.query_log, and metadata logs.Test plan
yarn startor CI docs check)Made with Cursor