fix: retry on Anthropic overloaded_error instead of halting session#20373
Open
OliPetry wants to merge 2 commits intoanomalyco:devfrom
Open
fix: retry on Anthropic overloaded_error instead of halting session#20373OliPetry wants to merge 2 commits intoanomalyco:devfrom
OliPetry wants to merge 2 commits intoanomalyco:devfrom
Conversation
Anthropic returns {"type":"overloaded_error","message":"Overloaded"}
as a stream error when at capacity. This was not recognized by
parseStreamError() (which only handled type=="error") or the fallback
JSON check in retryable(), causing the session to halt with a terminal
error instead of retrying with exponential backoff.
- Add overloaded_error handling to parseStreamError() as a retryable
api_error
- Widen ParsedStreamError.isRetryable from literal false to boolean
- Add overloaded_error check to retryable() fallback JSON path as
a safety net
github-actions
bot
added
needs:compliance
Contributor
|
Thanks for your contribution! This PR doesn't have a linked issue. All PRs must reference an existing issue. Please:
See CONTRIBUTING.md for details. |
github-actions
bot
removed
needs:compliance
Contributor
|
Thanks for updating your PR! It now meets our contributing guidelines. 👍 |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Issue for this PR
Closes #20384
Type of change
What does this PR do?
Anthropic returns
{"type":"overloaded_error","message":"Overloaded"}as a stream error when their API is at capacity. This was not recognized by the error handling pipeline, causing sessions to halt with a terminal error instead of retrying.The root cause:
parseStreamError()inerror.tsonly checked forbody.type === "error", so"overloaded_error"fell through. The error became aNamedError.Unknown, andretryable()inretry.tsdidn't match it either since its JSON fallback path only checked fortype === "error"variants.The fix adds
overloaded_errorhandling in two places:parseStreamError()— recognizestype: "overloaded_error"and returns it as a retryableapi_error. Also widensParsedStreamError.isRetryablefrom literalfalsetobooleanso stream errors can be retryable.retryable()fallback — addsjson.type === "overloaded_error"check as a safety net for the JSON parsing path.With this fix, overloaded errors trigger the existing exponential backoff retry logic (2s → 4s → 8s → 16s → 30s cap) instead of killing the session.
How did you verify your code works?
fromError()→parseStreamError()→retryable()confirming the classification bugparseStreamError()now returns{type: "api_error", isRetryable: true}, whichfromError()wraps asMessageV2.APIError, whichretryable()matches at line 54-58 (existingisRetryable+"Overloaded"check)Checklist