Repository navigation
Setting up Bazel and README - #2
Merged
Merged
Conversation
behinddwalls
marked this pull request as ready for review
January 14, 2026 21:35
Contributor
|
Collaborator
Author
|
djuloori2794
approved these changes
Jan 14, 2026
kevinlnew
added a commit
that referenced
this pull request
Jun 16, 2026
Scaffold the runway domain with entity types for the check and land queues, plus pipeline topic key constants. Runway reuses the shared entity/change and entity/mergestrategy packages rather than defining its own strategy type. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
kevinlnew
added a commit
that referenced
this pull request
Jun 16, 2026
Scaffold the runway domain with entity types for the check and land queues, plus pipeline topic key constants. Runway reuses the shared entity/change and entity/mergestrategy packages rather than defining its own strategy type. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This was referenced Oct 27, 2026
sbalabanov
added a commit
that referenced
this pull request
Oct 30, 2026
## Why? `process` admits the latest head with two ordered CAS writes on two rows: it increments the anonymous `queue.in_flight_count` first, then moves the request accepted→processing. A crash between them leaves the count raised with no owner. Holds never consume retry budget and the DLQ reconciler releases only `processing` requests, so the queue wedges permanently. 1. Queue at `max_concurrent = 1`; request R is accepted and is the latest head. 2. `process` claims the slot (`in_flight_count = 1`) and crashes before writing `processing`. 3. Every redelivery re-reads `accepted`, finds the gate closed by its own claim, and holds. 4. Nothing admits again, and the best-effort unwind releases never ran because the process crashed. ## What? - `admitLatestHead` gates and derives the strategy as before, then CASes the request accepted→processing first and increments `in_flight_count` second. A version conflict on the increment reloads and retries without re-gating or coalescing, because the admit is already durable. A lost request CAS acks without touching the counter. - Removes `process`'s `releaseBuildSlot` helper and both unwind call sites. - Trade-off: the count now drifts low rather than high. A crash or hard failure between the two writes, or two heads that both pass the gate in the same window, over-admits by one. The existing zero clamp in the `buildsignal` and DLQ release paths brings the count back to the true value as the queue drains. The count can go high only if a delivery stalls between the two writes past the visibility timeout and wakes after its redelivery has finished the validation; the RFC records this residual and how to narrow or close it. - `doc/rfc/stovepipe/steps/process.md`: step 7 reordered; the redelivery and `in_flight_count integrity` sections now describe the fail-open order. Two `dlq.go`/`buildsignal.go` doc comments that described the old claim-first order are corrected (comments only). - This change does not repair counts that leaked before it, so a queue that is already wedged needs a one-time manual reset of `in_flight_count`. Alternative to #966, which keys slots by request ID in a new `in_flight_request_ids` column and is the exact fix; this PR needs no schema change. Merge one or the other, not both. ## Test Plan - `make fmt && make lint && make check-tidy && make check-gazelle && make check-mocks` - `./tool/bazel test //stovepipe/... //service/stovepipe/... --test_output=errors` - `./tool/bazel build //test/e2e/stovepipe:go_default_test //test/integration/stovepipe/...` Audit: SP-OCT-003 and SP-OCT-005 in doc/audit/stovepipe-fable-v2.md (Sol v2 #1/#2, Opus v2 #1/#8). 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: sergeyb <sergeyb@uber.com> Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
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.
This change to set-up the Bazel build and test
Changes Include,