Httpx timeout in director-v0 bug #7456
Description
Is there an existing issue for this?
- I have searched the existing issues
Which deploy/s?
other (e.g. local)
Current Behavior
The director-v0 has the httpx client of fastAPI at default settings, with a default response timeout of 5sec. Failed calls are retried. This leads to weird behaviour when the docker registry takes >5sec, e.g. 10-15sec, to respons to a request, as the requests are canceled and retried by dv0 after 5sec. This overloads the registry.
The catalog has the httpx-timeout custom set to 20sec, which might also be appropriate for dv0?
Expected Behavior
No response
Steps To Reproduce
No response
Anything else?
No response