Skip to content

fix(zig): resume managed embedding backfill after restart - #472

Closed
ajroetker wants to merge 1 commit into
agent/ha-failover-runtime-hardeningfrom
codex/fix-managed-embeddings-restart-flake
Closed

fix(zig): resume managed embedding backfill after restart#472
ajroetker wants to merge 1 commit into
agent/ha-failover-runtime-hardeningfrom
codex/fix-managed-embeddings-restart-flake

Conversation

@ajroetker

Copy link
Copy Markdown
Contributor

Summary

  • restore metadata-owned managed embedding providers during startup catch-up
  • keep a short-lived startup probe from terminalizing transient provider failures at the persisted retry-budget boundary
  • add a focused unit regression and include it in the API table-write test gate

Testing

  • zig build -Dedition=full install -fincremental
  • focused startup catch-up unit regression: 1 passed
  • serial repository regression loop: 10/10 passed
  • concurrent repository regression loop: 3 workers x 5 iterations, 15/15 passed
  • git diff --check

Context

Follow-up for #347. Reproduced from https://github.com/antflydb/antfly/actions/runs/31531357544/job/93912435099?pr=347.

@ajroetker

Copy link
Copy Markdown
Contributor Author

Superseded by commit 96917f1 on #469, where the fix can run against the combined branch and its main-targeting CI.

@ajroetker ajroetker closed this Aug 11, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant