You are not logged in.

#1 2021-03-22 20:33:43

Registered: 2018-11-29
Posts: 5

Reasons for random system freezes

Hi all,

I'm pretty sure this is not a BL issue, so kinda off topic, I'm just fishing for hints here.

A couple of day ago my system started to completely freeze every now and then, usually when I play a game (probably because it puts the hardware under a lot of stress), but it has also happened when the system was basically idle.
As a result, the only way to get out of it is by cutting power. Can't even start a terminal session or log out anymore.

Now the problem is that I cannot run any diagnostics when everything is dead. I managed to get into the terminal once, where I was bombed with pretty weird I/O error messages when trying to type.

First idea for the culprit was the SSD. I actually found out that my model is known to cause random freezes after a while and required a firmware update, but that didn't solve the problem either.

Any ideas as to what else it could be? RAM? CPU/GPU, maybe an overheating issue? Don't really know where to start looking here, and a google search for freeze+debian naturally doesn't get me anywhere lol...



#2 2021-03-23 18:19:31

From: Saskatchewan, Canada
Registered: 2017-06-27
Posts: 79

Re: Reasons for random system freezes

Instead of cutting power, have you ever done 'SysRq'? (often accessed by alt-print screen) then press some keys. 
I use R E I S U B mostly. 

Link to wikipedia which has a table of keys you can press after doing SysRq.

Last edited by trilobite (2021-03-23 18:19:49)

{Linux-using people I haven't met are friends yet to be made.}


#3 2021-04-04 14:39:10

Registered: 2019-06-10
Posts: 26

Re: Reasons for random system freezes

Yep probably your SSD. I had the same problems on a failing USB, weird I/O error messages on the terminal. Those I/O error messages probably won't occur if you were to cd to a tmpfs or filesystem not on that SSD.

SysRq R E I S U B is good advice.

Some workarounds aside from getting a different SSD:

Putting /var/log on a tmpfs will allow you to be able to read logs leading up to the problem. That is if you have a terminal ready before the problem, on a virtual console (e.g. shift+ctrl+f6) possibly, as you can't launch one later as you've observed.

EDIT:Also many utilities to read the logs probably won't work, I think I used 'cat' or tail or other simple combinations, vi/nano wouldn't work. Also sudo wouldn't work either IIRC, the logged in tty user had to have permissions to read /var/log in the first place, e.g. a member of the 'adm' group.

In my /etc/fstab:

tmpfs /tmp tmpfs nosuid,nodev,noexec 0 0
tmpfs /var/log tmpfs nosuid,nodev,noexec 0 0

And to get around apt requiring /tmp to be exec, make /etc/apt/apt.conf.d/50remount:

DPkg::Pre-Install-Pkgs {"mount -o remount,exec /tmp";};
DPkg::Post-Invoke {"mount -o remount /tmp";};

Also consider moving cache folders (e.g. ~/.cache subdirectories) to a tmpfs appropriate to whatever software you use to reduce I/O load on your SSD.

Also, that USB drive I mentioned eventually failed less than a year after purchase, terrible design by SanDisk, was too hot to touch and I bet wasted a bunch of electricity in comparison to other models, a way overpriced resistor, a short or whatever... I wouldn't be surprised if that was the reason it failed. Many others complained. If you have an IR camera take a look. Something to consider.

P.S. I've found Toshiba, now KIOXIA to be very reliable... worth spending the extra $.

Last edited by AndrewSmart (2021-04-05 00:59:44)


Board footer

Powered by FluxBB