My server keeps having bouts of unresponsiveness. When it is unresponsive 50% of the time, it's just the GUI/SSH/Console that are unresponsive. VMs and containers keep working. Other times my containers are unresponsive also (Plex and others). When I say console, I mean directly connected to the server with KVM. When I try to login via KVM I enter the username and it hangs for 60 seconds and then says the login timed out. I finally hit "control alt delete" and the server started to reboot, and did a forced shutdown. Screenshot of those attached as well. I do know that I left a GUI link up to my Unraid server on my "sleeping" MacBook. The logs that I see when I'm able to get into the GUI are just generic php-fhm and www pool logs, that I find very little about other than "don't keep a GUI open" (which blows my mind that is still an issue). Please help. This diagnostic was generated automatically when doing the forced shutdown. 

 

 

dundermifflin-diagnostics-20241231-0833.zip syslog-previous

Dec 31 08:00:00 DUNDERMIFFLIN php-fpm[6674]: [WARNING] [pool www] child 6851 exited on signal 9 (SIGKILL) after 42.619479 seconds from start
Dec 31 08:00:02 DUNDERMIFFLIN php-fpm[6674]: [WARNING] [pool www] child 7645 exited on signal 9 (SIGKILL) after 23.544891 seconds from start

 

It's filled with these, try booting in safe mode and/or closing any browser windows open to the GUI, only open when you need to use it then close again, it can help with these errors.

It booted fine after rebooting, but eventually happens again. I’ll try to keep the browser closed, but I forget often. Happen to know if there’s an actual bug for this or a fix in the works? 

21 minutes ago, 62165 said:

Happen to know if there’s an actual bug

Not sure, it doesn't happen to all users, but keeping the GUI closed helps for some.

3 hours ago, JorgeB said:

Dec 31 08:00:00 DUNDERMIFFLIN php-fpm[6674]: [WARNING] [pool www] child 6851 exited on signal 9 (SIGKILL) after 42.619479 seconds from start
Dec 31 08:00:02 DUNDERMIFFLIN php-fpm[6674]: [WARNING] [pool www] child 7645 exited on signal 9 (SIGKILL) after 23.544891 seconds from start

 

It's filled with these, try booting in safe mode and/or closing any browser windows open to the GUI, only open when you need to use it then close again, it can help with these errors.

Crashing again, no GUI open. Can’t get in to see logs, but assuming the same thing. 

Another crash. Another set of logs/diagnostic...

 

 

syslog-previous dundermifflin-diagnostics-20241231-1651.zip

Try booting in safe mode to rule out any plugin issues.

I think I’ve gotten to the bottom of this. I was going to try safe mode, but I needed my plugin to pass through USB devices (didn’t want to reinvent the wheel). I booted in GUI mode and logged in. When the issue started happening, I was able to get the terminal up after many attempts and it took about 5 mins to load…then I was able to run HTOP. Ultimately, I built a new VM the other day and allocated it 16GB RAM and have 32GB in the system. Docker was using 18.6GB, ZFS was using 4GB, then my VMs were using 25GB more. Ultimately, I over-allocated memory. I was thinking it would share memory like CPU but I guess that’s not the case. Lowered RAM allocations and it has been stable without the php-fpm errors for several hours.