Workers are stuck at initializing and throttled states for more than 3 hours already.
This endpoint was deployed via importing a github repo, the first iteration started building its docker image immediately after deploying the endpoint.
However, I had to make changes to the python script, and I only modified the python script. After I pushed the changes, I canceled the deployment of the first iteration. Then waited for the automatic detection of the changes in the repo. Since then it has been stuck on initializing and throttled cycle. While in the builds tab, the build is still pending.
How do I fix this?
10 Replies
Same issue here. We have a serverless endpoint that has been running for a year that is now stuck initializing. No changes were made to it.
@Karolis how did you solve it? did you make another endpoint?
using a different repo?
We're using Runpod's Infinity Embeddings repo. Tried starting a new endpoint, tried different data centers, tried different GPUs - nothing works.
We haven't solved it. Filed a ticket with Runpod - no answer yet.
Correct me if im wrong, but that is a ready-made endpoint right?
Yes
Samee, happening with me too. Its annoying at this point
Same here something seems broken on RunPod’s side. It’s been a day with no updates or help from RunPod, this is sad
Unknown User•2w ago
Message Not Public
Sign In & Join Server To View
Same problem here
Unknown User•2w ago
Message Not Public
Sign In & Join Server To View