Forte Hung - just once #62

gemlog commented 2026年04月21日 17:20:09 +02:00

I'm hesitant to post this as an Issue, b/c it was weird.
I was happily using Forte for a couple of hours when it suddenly hung. Shift+Refresh was no help. A new container tab was no help. I checked a couple of other sites on the same VPS and they were up just fine. Nginx had been reloaded via cron about an hour earlier, but I did an actual restart anyhow. That made no difference. Then, before I could shell in and have a good look around, Forte began responding again. It's a fairly weak VPS, but I typically have 2GB free out of a total of 4GB and there's plenty of storage left - plus the other sites were responding normally.
I checked /var/log/forte/php.log, which is 105MB; too big to load into my editor. I tailed the last 500 lines to less, but I don't even know what I'd be looking for. I hope that gets rotated, I'll have to keep an eye on it (that's my problem though).
Does Forte have some maint. that it does?
Anyhow, it's working fine again now. IDK.

I'm hesitant to post this as an Issue, b/c it was weird. I was happily using Forte for a couple of hours when it suddenly hung. Shift+Refresh was no help. A new container tab was no help. I checked a couple of other sites on the same VPS and they were up just fine. Nginx had been reloaded via cron about an hour earlier, but I did an actual restart anyhow. That made no difference. Then, before I could shell in and have a good look around, Forte began responding again. It's a fairly weak VPS, but I typically have 2GB free out of a total of 4GB and there's plenty of storage left - plus the other sites were responding normally. I checked /var/log/forte/php.log, which is 105MB; too big to load into my editor. I tailed the last 500 lines to less, but I don't even know what I'd be looking for. I hope that gets rotated, I'll have to keep an eye on it (that's my problem though). Does Forte have some maint. that it does? Anyhow, it's working fine again now. IDK.

abanink commented 2026年04月21日 19:23:38 +02:00

It's always interesting to look for log entries just before, during and just after the time of the unavailability. If you don't have a clue and there is no confidential information in them, then you can share them here for us to have a look.

php.log could be interesting but also your nginx (access) log files if you have those configured. It could show which requests made it to the webserver at all, and what the response was if any.

I'm running forte for months now on multiple machines and I've never seen what you report, and I use it on a daily basis. One possible difference is that I self-host it from source on my dev server and also like this (install option 1) in a rootless container running nixos.

It's always interesting to look for log entries just before, during and just after the time of the unavailability. If you don't have a clue and there is no confidential information in them, then you can share them here for us to have a look. php.log could be interesting but also your nginx (access) log files if you have those configured. It could show which requests made it to the webserver at all, and what the response was if any. I'm running forte for months now on multiple machines and I've never seen what you report, and I use it on a daily basis. One possible difference is that I self-host it from source on my dev server and also like [this (install option 1)](https://codeberg.org/abanink/forte-nix) in a rootless container running nixos.

gemlog commented 2026年04月21日 21:49:05 +02:00

I have it. It just happened again, but I managed to shell in to see that peertube was eating both cpu's alive!
Actually, it was ffmpeg doing it. I don't even know what peertube is written in or it may be some config I can do somewhere. Maybe there's a way to 'nice' it. I'll look.
I knew it didn't sound right for an issue, but now I've caught the culprit in the act.
Yes, I would certainly share my log with you guys, if it was useful, but this time it is not: I will close this - I was just baffled this morning.

I have it. It just happened again, but I managed to shell in to see that peertube was eating both cpu's alive! Actually, it was ffmpeg doing it. I don't even know what peertube is written in or it may be some config I can do somewhere. Maybe there's a way to 'nice' it. I'll look. I knew it didn't sound right for an issue, but now I've caught the culprit in the act. Yes, I would certainly share my log with you guys, if it was useful, but this time it is not: I will close this - I was just baffled this morning.

gemlog commented 2026年04月21日 21:49:48 +02:00

Unrelated to the Forté project - it was an issue on my server.

Unrelated to the Forté project - it was an issue on my server.