System locking up. Which log to look at to see what's happening?

Greetings,
I have a Free BSD workstation that randomly is locking up and requires a hard reboot.
There does not seem to be a reason for this.
Turning on the monitor (I keep it powered off unless I'm working at this workstation) does not show anything, the
screen is black, and it requires a power-cycle to come back online.
Which log file might indicate what the issue is?
Thanks in advance.
T
 
you can check /var/log/messages
but if there's hardware issue like bad hard disk there will be no logs. You need to test the hardware.
 
Turning on the monitor (I keep it powered off unless I'm working at this workstation) does not show anything, the
screen is black, and it requires a power-cycle to come back online.
Maybe it's just the graphics that have been switched off (and fail to come back on). Can you access the system with ssh(1)?
 
Maybe it's just the graphics that have been switched off (and fail to come back on). Can you access the system with ssh(1)?
SirDice,
Thanks for taking the time to look at my post.
ssh will not connect and it doesn't respond to pings., and shows as offline on the Tailscale dashboard.

Sorry, I should have mentioned that in the original post.
T
 
ssh will not connect and it doesn't respond to pings
That's definitely a bad sign. It's going to be difficult to diagnose though. If it's this locked up it won't be able to log anything either. But, definitely check /var/log/messages, there may be clues leading up to the lock up.

Some general hardware advice, check memory, check fans (temperatures), check disks, etc. Try and rule out any hardware issues.

Oh, any power saving stuff enabled? Maybe the system went to sleep and refuses to get back online?
 
When the system boots, start a ping on another host on the same local network. And start a ssh login, in which you run something like top or "vmstat 1". That at least gives you an idea of the last thing that happened before the crash, and you can use that as a starting point.

You can also do something like "ps aux" or "vmstat 1" or "top" and save it in a local file, making sure it is not overwritten after the reboot. Again, that's not a guaranteed diagnosis, but a starting point for debugging.
 
It's not a i9 13900K? Mine has a random freeze once per month at high load. 😆
It doesn't look promising, but some things you could try is to get into the usual environment from a remote terminal. It will probably lose connection but might give a final hint. In addition to that, you can run a script that repeatedly sends particular status information to a remote system using an nfs-server or nc listen process. This takes soome manpage reading.
 
Similar hard lockup issue with FreeBSD 13.5-STABLE and a Beelink Ser 5 Pro MIni PC, AMD Ryzen 7 5825U (32G RAM, 1TB SSD). Locks up approx every 6 weeks. Video dead (monitor goes to sleep). ssh not working, but somewhat curiously pings continue to work. Nothing in any of the logs. Memory tested OK. A thorough workout rebuilding the OS with -j8 from source never has an issue (so not a high load issue). Perplexing!

Will try ralphbsz's solution to track down its final moments.
 
What's running usually? It sounds like a memory leak. Process growing over time until everything clogs by out of memory and swapspace.
 
What's running usually? It sounds like a memory leak. Process growing over time until everything clogs by out of memory and swapspace.
In my case, just xorg, twm, ipfw, bind (split DNS to resolve local network machines, other queries forwarded), Apache web server, mail server (sendmail), a few sendmail milters including my own one and SeaMonkey. Never had a memory leak problem when running the same selection on Intel Mac minis and my sendmail milter (written in C) runs on a university mail server (admittedly under Solaris rather than FreeBSD) without issue. System has now been up 1 day 10 hours, the 10G swap has not been touched and top shows 19G free memory but I'll monitor from a second system as suggested above.
 
Back
Top