Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Auto retry of the requests #88

Open
tomaszkiewicz opened this issue Dec 30, 2024 · 3 comments · May be fixed by #91
Open

Auto retry of the requests #88

tomaszkiewicz opened this issue Dec 30, 2024 · 3 comments · May be fixed by #91
Labels
enhancement New feature or request

Comments

@tomaszkiewicz
Copy link

Is your feature request related to a problem? Please describe.

When hitting throttling limits, the gateway throws errors causing end clients to wait for reply for a long time, and the reply never comes.

image

Above shows that the gateway are able to catch that case (ThrottlingException) and that underlying libraries have some ability to set retries.

So this feature request is about:

  • either exposing number of retries to the ENV variables so we can set them (probably with some backoff strategy)
  • passing the proper errors to clients, so they can fail fast

With #1 being a preferred option :)
@tomaszkiewicz tomaszkiewicz added the enhancement New feature or request label Dec 30, 2024
@tbtommyb
Copy link

tbtommyb commented Jan 6, 2025

It would help if these errors were returned at 429s so that consuming libraries could implement their own retry logic. Should _invoke_bedrock check for ThrottlingException rather than returning a 500 for all errors?

@Wh1isper
Copy link

Wh1isper commented Jan 7, 2025

Given that the most clients have implemented retry, I prefer to return a 429 error

@Wh1isper Wh1isper linked a pull request Jan 9, 2025 that will close this issue
Open
@Wh1isper
Copy link

Wh1isper commented Jan 9, 2025

I just created a PR for this: #91. Not sure if I missed anything. Thx!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
enhancement New feature or request
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

feat: Response 429 if ThrottlingException Wh1isper/bedrock-access-gateway
3 participants
@tomaszkiewicz @tbtommyb @Wh1isper