Tagger error: 502 - Bad Gateway

[portal-surfacing]

Hi,
I do not know how docker related this issue is, at least I am running the stashapp/stash container.
The error {"networkErrors":{"code":502,"message":"Response body Bad Gateway"},"graphqlErrors":null} appears when running Scrape by fragment in the Tagger view. Before the error pops up, it loads for a while (~ 30 seconds). But the Testing Credentials in the Metadata Providers settings show a Successful authentication just <1 second after klicking the Test Credentials button. That is also why I think there might be no issue on my docker host setup, but more likely within some docker configuration?

I do not now how to root the cause, my still running old instance is working fine.
Running an unprivileged debian container in proxmox with docker from their own repositories installed. My old, still running stash instance is with the almost same docker compose yaml on a debian VM on another proxmox instance but not sure if this makes a difference. Always on wireguard VPN with killswitch and allowance of local LAN traffic (for old and new instance). Also tested with a VPN from a different provider, did not help.

Any help/hints welcome. Maybe I should try out another container? Just found out about with the nice Container Comparison Matrix here on discourse.

Thank you very much and best.

[WithoutPants]

A 5xx error indicates a server error. It seems strange that the authentication would work but the query does not. Does the query search do the same thing?

It seems unlikely that the 5xx error is coming from the stash box instance. Do you have a proxy or something that your connection goes through?

[portal-surfacing]

Fixed it, the two different tested VPNs are not working for tagging. Trying out another one, the one from the old working stash instance, which in fact is from a third provider, worked for the new instance as well. And vice versa, the not working VPNs led to the 502 error on the old stash instance.

Before figuring this out, I tried the following things unsuccessfully:

1. Changing the docker from official stash to feederbox826’s stash-s6
2. Changing the proxmox network interface from masquerading to a default bridge (Network Configuration - Proxmox VE)

Another idea was to inspect with traceroute, but that was not helpful. At least I got to knew a few big hosters which seem to be commonly used in the VPN sector

With the broken VPN, when doing query search, the same error happened. But it just showed up in the info log in contrast to Scrape by fragment. There, the error is additionally written in red on the tile entry that is currently being searched in the tagger view. In the info log the error is also a bit more specific: scrapeSingleScene: input: scrapeSingleScene {"networkErrors":{"code":502,"message":"Response body Bad Gateway"},"graphqlErrors":null}