Repository navigation
Fail KV workloads on final errors and wrong data - #69
Draft
KirillKurdyukov wants to merge 1 commit into
Draft
KirillKurdyukov wants to merge 1 commit into
KirillKurdyukov wants to merge 1 commit into
Conversation
This was referenced Oct 15, 2026
KirillKurdyukov
marked this pull request as draft
October 15, 2026 09:02
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Changes
workloads on an unrecovered final logical operation error, preserving its
SDK cause through supervision and cleanup.
and full payload. Check successful reads before advancing and after each
acknowledged write, including during ordinary load.
table for the complete create/run/drop lifecycle.
Preserve existing schema, CLI, retry policies and load metrics.
Verification
The runner regression exercises stale-data detection while load is active.
same-label peers keep separate tables and the longer peer survives the
shorter peer's cleanup.
generation, wrong payload and a final SDK failure for each adapter.
Data injection follows actual successful SDK read/decoding; SDK failures
come from an independently dropped acknowledged private table.
Both SDK roles currently use the same SDK library revision, so this is
adapter/ownership verification rather than a two-version benchmark.
Recovered successful outcomes are covered at the core seam; native forced
retry recovery and hosted SLO remain unverified.
Consumer dependency
The separate
ydb-java-sdkworkflow PR must pin these changed examples at68c06dfa5f09dea42613dc5eaffb0e4c456cfd51and enforce the process verdict.That consumer publication/CI is a follow-up, not an unchanged-master success
claim. This PR adds neither RMW nor Topic scenarios and changes no SDK library.