Skip to content

Version 1.7.3 is completely broken: High amount of failed and retried requests #368

@ashleykleynhans

Description

@ashleykleynhans

Describe the bug
After upgrading to version 1.7.3, the number of retried and failed requests spiked dramatically - Of 155 total requests, 17 were retried and 11 failed.

This is not happening with version 1.7.2, and things are back to normal after reverting to version 1.7.2.

To Reproduce
Steps to reproduce the behavior:

  1. Use a ComfyUI serverless worker.
  2. Ensure that runpod python module is version 1.7.3.
  3. Send a few dozen requests to the endpoint.
  4. Observe a high number of retried and failed requests.

Expected behavior
There should be NO retried or failed requests.

Screenshots
Screenshot 2024-10-16 at 08 33 39

Desktop (please complete the following information):
N/A - Specific to RunPod Serverless.

Smartphone (please complete the following information):
N/A - Specific to RunPod Serverless.

Additional context
None.

Metadata

Metadata

Assignees

Labels

bugSomething isn't workingfixed

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions