mirror of
https://github.com/fosrl/pangolin.git
synced 2026-07-16 14:47:04 -05:00
[GH-ISSUE #1090] 502 errors after upgrading to pangolin 1.7.x #27376
No Assignees
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: github-starred/pangolin#27376
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @L0sWach0s on GitHub (Jul 18, 2025).
Original GitHub issue: https://github.com/fosrl/pangolin/issues/1090
Originally assigned to: @oschwartz10612 on GitHub.
Hello everyone!
I upgraded my Pangolin instance to 1.7.x today. The 1.6.2 was working cleanly so far.
Unfortunately, since the update to 1.7.x (regardless of which version, I've tried them all), I'm constantly getting a 502 error from Cloudflare (proxy is active) on my sites. This also means I can no longer use my Authentik instance.
A downgrade (thanks to backup) to 1.6.2 without changing my settings solved the problem.
snip of the log:
{"level":"debug","time":"2025-07-18T19:08:55+02:00","caller":"github.com/traefik/traefik/v3/pkg/server/service/loadbalancer/wrr/wrr.go:175","message":"Service selected by WRR: https://100.89.128.4:53393"}/ {"plugin":"plugin-badger","module":"github.com/fosrl/badger","runtime":"","time":"2025-07-18T19:08:55+02:00","caller":"fmt/print.go:305","level":"debug","message":"Badger: Valid session"} {"level":"debug","time":"2025-07-18T19:08:55+02:00","caller":"github.com/traefik/traefik/v3/pkg/server/service/loadbalancer/wrr/wrr.go:175","message":"Service selected by WRR: https://100.89.128.4:53393"}/ {"plugin":"plugin-badger","module":"github.com/fosrl/badger","runtime":"","time":"2025-07-18T19:08:56+02:00","caller":"fmt/print.go:305","level":"debug","message":"Badger: Valid session"} {"level":"debug","time":"2025-07-18T19:08:56+02:00","caller":"github.com/traefik/traefik/v3/pkg/server/service/loadbalancer/wrr/wrr.go:175","message":"Service selected by WRR: https://100.89.128.4:53393"}/ {"plugin":"plugin-badger","module":"github.com/fosrl/badger","runtime":"","time":"2025-07-18T19:09:02+02:00","caller":"fmt/print.go:305","level":"debug","message":"Badger: Valid session"} {"level":"debug","time":"2025-07-18T19:09:02+02:00","caller":"github.com/traefik/traefik/v3/pkg/server/service/loadbalancer/wrr/wrr.go:175","message":"Service selected by WRR: https://100.89.128.4:53393"}/ {"level":"debug","error":"read tcp 100.89.128.1:38618->100.89.128.4:53393: read: connection reset by peer","time":"2025-07-18T19:09:02+02:00","caller":"github.com/traefik/traefik/v3/pkg/proxy/httputil/proxy.go:121","message":"502 Bad Gateway"}@oschwartz10612 commented on GitHub (Jul 18, 2025):
Hi @L0sWach0s sorry about this issue.
Could you confirm if when you are upgraded if Newt is connected properly? It feels like it is not able to address your resource on the other end but that could be a myriad of things.
@L0sWach0s commented on GitHub (Jul 18, 2025):
Hey @oschwartz10612!
Newt was connected properly and it was also the latest version. I could reach out to the containers, but often I got the 502 of Cloudflare. I tried different clients, even some who newer got in touch with my domain.
@joediont commented on GitHub (Jul 19, 2025):
I have also just upgraded to Pangolin 1.7.2 this week and was having a lot of intermittent issues with 502 errors and resources like immich lagging and really slow performance.
What fixed my issue was upgrading my newt instance on my Truenas which was still on 1.3.2. Truenas native app reports Newt as latest even though the latest Newt should be 1.3.4. I used portainer to deploy the latest instance and everything is back to stable as before the upgrade.
Hopefully it's the same issue, otherwise good luck with the debugging!
@L0sWach0s commented on GitHub (Jul 19, 2025):
I checked it and the newt container, who hosted the Authentik site is on 1.3.4...
newt -versionNewt version 1.3.4@cfurrow commented on GitHub (Jul 20, 2025):
I have also noted a large uptick in 502 errors from some resources when I updated to Pangolin 1.7.3 (I was on 1.2.x). I also updated newt from 1.3.0->1.3.4.
When looking at the docker logs for newt and pangolin, there was a lot of noise and repeated messages:
newt logs
Pangolin also had repeating logs:
Every so often I'd see log lines like this:
I had not setup newt docker socket access, so that was expected.
I did notice that I had a docker hostname being used in one of my resources, and I'm not sure how that was working prior to newt having docker socket access.
I updated my docker-compose.yml per newt's README.md instructions, recreated the newt container, and the logs stopped being so noisy
I'm writing this at 11:35am ET, and Pangolin has not had new log output since 11:20am, versus repeating logs every second or so.
Newt is similarly quiet, with no new log output since 11:20am ET as well, which is when I recreated the newt container:
@L0sWach0s commented on GitHub (Jul 20, 2025):
THIS!
@cfurrow: You gave the right hint!
When I deactivated the Docker integration settings for the site (because I didn't allow the usage of the Docker socket on the newt client), the issue was resolved!
I thought the server would handle it correctly, even when the client wasn't sending information from the Docker socket.
I will continue to monitor this behavior and then close the ticket.
Regards
@oschwartz10612 commented on GitHub (Jul 21, 2025):
Hum... Very good info thanks for this. I will look into it ASAP and see if I can tell why the docker socket would be messing with connectivity. Thanks for the info!
@schiggieh commented on GitHub (Jul 22, 2025):
Thanks, did the trick so far!