SGPINF-961: fix: add litellm retry with exponential backoff for rate limit errors#257
Merged
sayakmaity merged 1 commit intomainfrom Feb 6, 2026
Merged
Conversation
- Add `num_retries=3` default to LLMConfig so litellm retries on OpenAI 429 rate limit errors with built-in exponential backoff - Increase Temporal DEFAULT_RETRY_POLICY from 1 attempt (no retries) to 3 attempts with exponential backoff (1s, 2s, 4s... up to 30s) This complements the HTTPX connection limit reduction in agentex backend (scaleapi/scale-agentex#144) to address OpenAI rate limiting under high concurrent load.
smoreinis
approved these changes
Feb 6, 2026
sayakmaity
changed the title
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
num_retries=3default toLLMConfigso litellm automatically retries OpenAI 429 rate limit errors with built-in exponential backoffDEFAULT_RETRY_POLICYfrommaximum_attempts=1(zero retries) tomaximum_attempts=3with exponential backoff (1s → 2s → 4s, max 30s)Context
Under high concurrent load (500 users × 40 messages), agent pods forward all requests to OpenAI simultaneously with no retry logic. When OpenAI rate limits kick in (429), the SDK immediately surfaces
ServiceUnavailableErrorwith no retry attempt.This PR is the SDK-side companion to scaleapi/scale-agentex#144, which reduces HTTPX connection pool limits in the agentex backend to add backpressure.
Together these two PRs:
Test plan
pytest tests/passes (141 passed, 2 skipped)ServiceUnavailableErrorcount drops significantly