handle 5xx by requeueing the request #9

Closed
opened 2023-07-29 22:37:07 +02:00 by Wroclaw · 3 comments
Owner

sometimes the openai api responses with 503, they ask to repeat the request. We should requeue the request instead of erroring the request.

sometimes the openai api responses with 503, they ask to repeat the request. We should requeue the request instead of *erroring* the request.
Wroclaw added the
bug
enhancement
labels 2023-07-29 22:37:07 +02:00
Author
Owner

it should be requeued to the front of the queue

it should be requeued to the front of the queue
Author
Owner

the same should go with 500
possible all 5xx (server side) errors

the same should go with 500 possible all 5xx (server side) errors
Author
Owner

if we handle those, we should also make sure to not make an accidental loop of it, so like 3 retries at max

if we handle those, we should also make sure to not make an accidental loop of it, so like 3 retries at max
Wroclaw changed title from handle 503 by requeueing the request to handle 5xx by requeueing the request 2023-07-31 20:12:13 +02:00
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference: Wroclaw/GPTcord#9
No description provided.