Repository navigation
Bump Temporal server ref and update repeated-update link expectation - #558
Merged
Merged
Conversation
Bump the dev server used by integration tests to temporal@de3cfe9 (current main). Among other changes, this picks up temporalio/temporal#11925 ("Fix links for repeated updates"). A Nexus Update operation that repeats an already-completed Update ID is not written to the target's history, so before #11925 its link pointed at a request ID with no event behind it. The server now links it to the original Update's accepted event by event ID instead. Update the NexusOperation/Sync/Update history matcher to expect that eventRef link. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
S15
approved these changes
Nov 17, 2026
Contributor
Author
|
So the build is failing because of a breaking change in the Rust SDK? From Claude: It seems strange to use a pinned release in some contexts, but not others. I might open another PR to fix that and get that squared away. |
… version sdk-rust main contains an unreleased breaking change to the Nexus operation start error type (temporalio/sdk-rust#1660), which breaks building the omes Rust worker from source. The rest of CI builds against the exact temporalio-sdk version pinned in workers/rust/Cargo.toml, so check out the matching tag here instead of the default branch. Other SDKs are unchanged. Co-Authored-By: Claude Opus 5.5 <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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What changed?
mise.toml_.server.ref) from temporal@0425053 (Aug 20) to temporal@de3cfe9 (currentmain).TestKitchenSink/NexusOperation/Sync/Updatehistory matcher. The repeated Update'sNexusOperationCompletedlink is now expected to be aneventReftoEVENT_TYPE_WORKFLOW_EXECUTION_UPDATE_ACCEPTED, instead of arequestIdRef.Why?
The bump picks up temporalio/temporal#11925 ("Fix links for repeated updates"). That PR intentionally changed this behavior.
The test runs a Nexus Update, then repeats the same Update ID with a new request ID. The repeat is never written to the target's history. Before #11925, its link referenced its own request ID, so it pointed at an event that doesn't exist. The server now links to the original Update's accepted event by event ID. The old expectation was written against the pre-#11925 server, so it asserted the dangling link.
I confirmed the cause by running the test against temporal@763c820. That commit is just before #12956 ("Sanitize Nexus return-path links") and already includes #11925. It fails the same way and records the same
eventReflink, so #12956 isn't involved.Getting past #11925 is also a prerequisite for testing newer server features in Omes, such as worker callbacks on standalone Nexus operations.
How did you test it?
Locally against temporal@de3cfe9:
TestKitchenSink/NexusOperation/Sync/Update/gopasses. It failed before the matcher change.scenariosandloadgen/kitchensink, andworkers/go.loadgen, the Go, Java, and PythonTestKitchenSinkcases pass.cargo build, proto-gen, or because the .NET runtime is missing, not on server behavior. CI covers those languages.Potential risks
This is a jump of about three months of server changes, so other integration tests could start behaving differently in CI. The local Go, Java, and Python runs found no other differences.
🤖 Generated with Claude Code