positron icon indicating copy to clipboard operation
positron copied to clipboard

Claude 4 Sonnet Bedrock throttling error

Open timtmok opened this issue 7 months ago • 1 comments

Describe the issue:

{
	"name": "ThrottlingException",
	"$fault": "client",
	"$metadata": {
		"httpStatusCode": 429,
		"requestId": "<REQUEST_ID>",
		"attempts": 3,
		"totalRetryDelay": 787
	}
}

Using Claude 4 Sonnet on AWS Bedrock returns this error when chatting with the model. This occurs on Positron desktop.

Steps to reproduce the issue:

  1. Sign in to AWS Bedrock
  2. Select Claude 4 Sonnet Bedrock in Assistant
  3. Send a message to the model

Expected or desired behavior:

It returns a response

Were there any error messages in the UI, Output panel, or Developer Tools console?

timtmok avatar May 26 '25 15:05 timtmok

This might be a configuration problem in AWS. When trying to use the model in AWS from the Chat/Text Playground, there is no throttling error.

https://repost.aws/knowledge-center/bedrock-throttling-error

timtmok avatar May 26 '25 15:05 timtmok

We didn't see the rate limiting error with Claude 4 Opus in #7843, but Heath is seeing this same rate limiting error for both Sonnet and Opus Claude 4 models:

The Claude 4 models work with the ellmer package from the R console, and Claude 3.7 works from the Positron Assistant.

Positron Version: 2025.06.0 (Universal) build 156
Code - OSS Version: 1.100.0
Commit: 5f45c9981a57f674e2be2676a4ae037ded3f8395
Date: 2025-05-28T03:34:50.368Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Darwin arm64 24.5.0

sharon-wang avatar May 29 '25 17:05 sharon-wang

To follow up on Sharon's post above and based on a suggestion provided in Tim's post above, I updated the AWS CLI to the latest version and still see the same issue.

--Thanks, all

heath-posit avatar May 29 '25 18:05 heath-posit

Sending this to QA as I think we haven't seen this since implementing https://github.com/posit-dev/positron/pull/8891 and https://github.com/posit-dev/positron/pull/8903, along with increasing our organizational token allowance.

Added Jonathan as assignee since he was the author of those PRs.

sharon-wang avatar Oct 07 '25 20:10 sharon-wang

Verified Fixed

Positron Version(s) : Positron Version: 2025.08.1 build 11 (palm)
OS Version          :  Palm (ubuntu?)

Test scenario(s)

Basic bedrock interactions not leading to any throttling errors.

Link(s) to TestRail test cases run or created:

testlabauto avatar Oct 09 '25 14:10 testlabauto