Skip to content

[REMOTE-1709] Back off harness auth secret fetch retries#11089

Open
abhishekp106 wants to merge 2 commits into
masterfrom
oz-agent/REMOTE-1709-auth-secrets-backoff
Open

[REMOTE-1709] Back off harness auth secret fetch retries#11089
abhishekp106 wants to merge 2 commits into
masterfrom
oz-agent/REMOTE-1709-auth-secrets-backoff

Conversation

@abhishekp106
Copy link
Copy Markdown
Contributor

@abhishekp106 abhishekp106 commented May 16, 2026

Description

  • Adds bounded retry with the shared out-of-band request retry strategy for list_harness_auth_secrets.
  • Keeps auth secret fetch state as Loading while retry attempts are pending, avoiding intermediate failure events.
  • Adds a 60s per-harness cooldown after final failure so picker repopulation/subscriber loops do not immediately start another retry chain while the server is down.

Linked Issue

REMOTE-1709

Testing

  • cargo fmt -- app/src/ai/harness_availability.rs

  • cargo check -p warp

  • cargo clippy -p warp --all-targets

  • I have manually tested my changes locally with ./script/run

works locally without spamming the server anymore!

Agent Mode

  • Warp Agent Mode - This PR was created via Warp's AI Agent Mode

CHANGELOG-NONE

This PR was created by Oz (running Codex).

Use the shared out-of-band retry strategy when fetching harness auth secrets and keep the model in Loading while bounded retries are pending.

After final failure, apply a short cooldown before ensure_auth_secrets_fetched can start another retry chain, preventing subscriber repopulation loops from spamming requests while the server is down.

Co-Authored-By: Oz <oz-agent@warp.dev>
@cla-bot cla-bot Bot added the cla-signed label May 16, 2026
@abhishekp106 abhishekp106 marked this pull request as ready for review May 16, 2026 01:39
@oz-for-oss
Copy link
Copy Markdown
Contributor

oz-for-oss Bot commented May 16, 2026

@abhishekp106

I'm starting a first review of this pull request.

You can view the conversation on Warp.

I completed the review and no human review was requested for this pull request.

Comment /oz-review on this pull request to retrigger a review (up to 3 times on the same pull request).

Powered by Oz

@abhishekp106 abhishekp106 requested a review from liliwilson May 16, 2026 01:39
Copy link
Copy Markdown
Contributor

@oz-for-oss oz-for-oss Bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Overview

This PR adds retry/backoff handling around harness auth secret fetches and keeps the UI in a loading state while retry attempts are pending.

Concerns

  • The retry predicate used for this GraphQL request does not classify GraphQL HTTP/status errors, so permanent failures such as 401/403/404 or user-facing GraphQL errors still burn through the retry chain and then enter the cooldown instead of failing fast.

Verdict

Found: 0 critical, 1 important, 0 suggestions

Request changes

Comment /oz-review on this pull request to retrigger a review (up to 3 times on the same pull request).

Powered by Oz

Comment thread app/src/ai/harness_availability.rs Outdated
Use a GraphQL-aware transient error classifier for harness auth-secret fetch retries so permanent GraphQL statuses and user-facing errors fail without consuming the retry chain.

Co-Authored-By: Oz <oz-agent@warp.dev>

Co-Authored-By: Abhishek Pandya <40919306+abhishekp106@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants