fix(llm): stop classifying Bedrock throttling as context overflow - #39623
Open
skogaby wants to merge 2 commits into
Open
fix(llm): stop classifying Bedrock throttling as context overflow#39623skogaby wants to merge 2 commits into
skogaby wants to merge 2 commits into
Conversation
Bedrock's ThrottlingException message "Too many tokens, please wait before trying again." matched the /too many tokens/i overflow pattern while the throttling exclusion only matched a literal "throttling error:" prefix the AI SDK never emits. The 429 was converted into a ContextOverflowError, which skipped retry and triggered a spurious auto-compaction instead of backing off. Broaden the exclusion list to match throttle vocabulary and transient retry language anywhere in the message, and gate parseAPICallError so a 429 status is never classified as context overflow.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 #39620
Type of change
What does this PR do?
Bedrock's ThrottlingException message "Too many tokens, please wait before trying again." matched the
/too many tokens/ioverflow pattern while the throttling exclusion only matched a literal "throttling error:" prefix the AI SDK never emits. The 429 was converted into a ContextOverflowError, which skipped retry and triggered a spurious auto-compaction instead of backing off.Broaden the exclusion list to match throttle vocabulary and transient retry language anywhere in the message, and gate parseAPICallError so a 429 status is never classified as context overflow.
How did you verify your code works?
Built a local copy of
opencodeand ran it on an account experiencing heavy throttling on AWS Bedrock. Normally this would trigger an auto-compaction almost immediately, but this time I saw proper backoff-retries.Checklist
If you do not follow this template your PR will be automatically rejected.