Repository navigation
fix(libsy): retain model call spans through completion - #915
Conversation
Signed-off-by: nachiketb <nachiketb@nvidia.com>
|
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configuration
📒 Files selected for processing (3)
Included review availability: This review used your included allowance. Your plan provides up to 12 included reviews per hour; 11 remain after this review. WalkthroughCallModel retains the originating tracing span. It records response usage and call outcomes on that span. The observability test checks that an abandoned run records an error outcome for the selected model. ChangesCall Span Telemetry
Priority: ➖ Normal Estimated code review effort: 2 (Simple) | ~10 minutes Merge Risk: ⚪ Minimal · up to The change records telemetry for completed and abandoned model calls. No actionable merge-blocking risk was identified. 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
I’m a rabbit, ears held high, Comment |
Why
Cancelling an algorithm while its model call is outstanding can leave
libsy.llm_callwithout an outcome. The outcome was recorded by the waiting algorithm, which may never resume.How
CallModel, matchingCallDecision.Validation
Extended one existing cancellation test with an outcome assertion. It fails before the fix (
Noneinstead oferror) and passes afterward. All 18 tests in the LLM client observability integration suite pass.Based directly on
main; independent of the Jev configuration MR.Summary by CodeRabbit