Pain point · Self-hosting & upgrades · Causes churn
Single Docker image fails to start after automatic update
6 source threads · first seen 2022-01 · last seen 2026-02
Summary
After Watchtower updated the single image, the container looped with an internal LiteLLM postgres timeout and returned 502; restart, re-pull and downgrade did not help.
- Affects
- self-hosted user running Docker single image
- Workaround
- Tried restarting, re-pulling, and downgrading without success
Evidence
Excerpts are copied word for word from the source; follow the link to read it in full.
“I tried restarting the container and re-pulling the image. A downgrade also did not work”
“After updating to 2.0.37 I can no longer add, delete, or move components on a few of my screens”
“Since upgrade to version 1.3.1 I am unable to open a screen in the app builder”
“so I'm stuck with an app that I cannot seem to make publicly accessible”
“got locked out of Budibase - all log in requests were returing in a 502 Time out”
“When I updated existing apps to the 1.0.43 version, then publish the app, it displays”