Repository navigation
Reduce product event collection and query overhead - #29
Merged
Merged
Conversation
|
Important Review skippedAuto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the
|
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Cortex product collection performs unnecessary learning reads and stores repeated event fields; log lists also fetch full detail payloads, and partial export failures replay already completed Queue messages.
Skip per-owner reads when no learning references exist, expose a bounded
analytics.eventListprojection, compact track/goal payloads while reconstructing the existing detail contract, and acknowledge messages only after all their raw and dictionary rows export. Mixed storage versions and cursor/dedup semantics remain supported.Validation: build, TypeScript, core (172 tests), Workers (117 tests including packed-message failure coverage), generated API docs, package consumer/dry-run checks, and maintenance unittest (3 tests) passed. Reproducible local benchmark: 100-user empty-reference SQL calls 205→106; sample pageview bytes 1093→808 and reference-goal bytes 1274→870; injected retry deliveries 14882→10000. These are synthetic counters/wire sizes, not production billing measurements.
Two independent parallel reviewers found no correctness issues. Added the suggested packed-message export-failure coverage and verified it. Schema-v2 writes require a compatible reader on rollback; deploy this runtime before Cortex adopts the new RPC.