System becomes unresponsive if left idle
 

Current stable Debian becomes unresponsive if left idle sometimes. Sometimes after one day, sometimes after 3 days.

Mouse cursor moves on black screen but system does not respond to ping, ssh or keyboard commands. No error I can discern in logs. It just freezes.

Any ideas how to diagnose and correct?

Thanks

I guess I would try memtest, check disks, PSU. Probably I would run some monitoring tool continuously.

How much free space do you have in your root partition?

How much free memory? Does it get used up over time?

Check the syslog file for clues but it is possible you are running some process that has a memory leak.

I don't appear to have any storage or memory issues at this moment. Barely scratching my reserves.

I do tend to leave my computer on for weeks or even months so it could happen slowly over time.

Will set up a cron job to log memory usuage, see what happens over time.

Thanks for the ideas.

Usually you can catch something in the logs right before the crash (like kernel panic) or look for coredumps. But in your special case probably I would implement a health check script to see if network, disk, ram, temperature, some processes are ok. And also would be nice to know if the kernel is still alive. Did you check the logs of X too?

Yeah I checked pretty much every log. Nothing of interest.

Did a BIOS update, just in case.

Included CPU, GPU, and memory stats into the monitor job already.

I'm about to go away so this computer will have until 4th Jan from today. Hopefully will recreate the issue and get some meaningful data.

Is this a "power management" issue? I've had struggles with systems not fully suspending/sleeping or resuming.

Is the system suspended? and not resuming? had it fully completed suspend? does it had during resume? Or are these disabled?

Came back to the machine today and it's stable, for the moment.

I'll update this thread when it happens again.

Well, it's still up and the problem seems to have vanished. ¯\_(ツ)_/¯

I've restarted once for a kernel update, but that's it. It's ticking along just fine.

Thanks for your input everyone. It's typical that the issue would magically disappear once I asked for help.