Skip to content

fix: queue concurrent streaming requests to avoid ~60s delay#12

Open
BP602 wants to merge 1 commit intorynfar:mainfrom
BP602:fix/queue-concurrent-requests
Open

fix: queue concurrent streaming requests to avoid ~60s delay#12
BP602 wants to merge 1 commit intorynfar:mainfrom
BP602:fix/queue-concurrent-requests

Conversation

@BP602
Copy link

@BP602 BP602 commented Feb 5, 2026

Summary

  • serialize Claude Agent SDK requests with p-queue to avoid the ~60s delay when concurrent streaming requests are made
  • add p-queue dependency and route all /messages requests through the queue
  • keeps streaming behavior intact while ensuring sequential execution

Testing

  • concurrent curl streaming requests (2 parallel requests) => ~12s total vs ~67s previously

Concurrent streaming requests to Claude Agent SDK would wait ~60s after
the first request completes before starting the second. This causes
significant delays when OpenCode makes concurrent requests (e.g., one for
the actual response and one for title generation).
Fix: Add p-queue to serialize requests (concurrency: 1), which eliminates
the 60s delay entirely. Requests now execute sequentially.
Before: ~67s total for 2 concurrent requests (4s + 60s delay + 3s)
After: ~12s total for 2 concurrent requests (4s + 8s sequential execution)
@BP602
Copy link
Author

BP602 commented Feb 5, 2026

Fixes #11

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