Intermittent crashing on Linux

I have been daily driving Omarchy for 253 days now and CachyOS for 182. I use Omarchy on my gaming rig and CachyOS for coding, spreadsheets and general admin tasks - don’t ask me why, it just happened this way.

Random Freezing

I have had some issues with Omarchy here and there, but nothing that I couldn’t solve. But recently I have been getting freezes that were unrecoverable where I had to physically go into the machine and push the reset button. It was alarming the first time it happened but I was lazy so I didn’t do anything about it.

But alas, it happened frequently enough that I had to do something:

$ journalctl --list-boots
IDX BOOT ID                          FIRST ENTRY                  LAST ENTRY
 -6 be8fce0c101447739acc51868266ca5a Wed 2026-08-05 18:58:25 AEST Wed 2026-08-12 11:41:13 AEST
 -5 ac21247913e94ca2ae65958a2539a4e4 Wed 2026-08-12 11:42:54 AEST Sat 2026-08-22 15:46:30 AEST
 -4 1ed6e5014c3e4639a91cf7893df95dc5 Sat 2026-08-22 15:50:06 AEST Sat 2026-08-22 15:53:21 AEST
 -3 cae5480e040f4f0389c717d73d5401fc Sat 2026-08-22 15:55:10 AEST Sun 2026-08-23 19:34:30 AEST
 -2 05996ecaebdd4a50b895e68cbbfb6a5e Sun 2026-08-23 19:38:14 AEST Thu 2026-08-27 00:32:22 AEST
 -1 b62bd0b0931749a3b306897f52f1a316 Thu 2026-08-27 02:19:20 AEST Fri 2026-08-28 20:53:10 AEST
  0 5dc084e810ec41a7a296b11965e58c0e Fri 2026-08-28 21:29:32 AEST Tue 2026-09-01 20:35:43 AEST

This machine is always on and those gaps show how long before I realized it was frozen and rebooted the machine.

Anyway, after faffing around the logs for a bit, I found these:

Aug 28 20:17:32 kernel: lazygit invoked oom-killer: gfp_mask=0x140cca(...), order=0, oom_score_adj=0
Aug 28 20:19:36 kernel: Node 0 active_anon:46001740kB inactive_anon:14529168kB ... anon_thp:31852544kB ... all_unreclaimable? yes
Aug 28 20:19:36 kernel: Node 0 Normal free:235900kB boost:172032kB min:237920kB low:301068kB high:364216kB
Aug 28 20:19:36 kernel: Free swap  = 160kB
Aug 28 20:19:36 kernel: Total swap = 4194300kB
Aug 28 20:21:55 kernel: Out of memory: Killed process 997894 (steamwebhelper) total-vm:52082328kB, anon-rss:459120kB ... oom_score_adj:300

Which I thought was weird, because I had 64GBs on this machine, then I remembered I added 2 16GB VMs around a week ago and suddenly it made sense (I thought). I then reduced the VM sizes to 10GB each and installed earlyoom as advised by Claude.

Random App Crashes

Thanks to that, my computer has not frozen yet. However, I have been seeing some weird behaviour with electron apps in particular, sometimes they would just stop working and would just show the shell. I also had an issue with steam where it would close and open in a loop, and then I realized that earlyoom has been killing my processes.

So one gotcha that I didn’t see in the logs originally was that the swap file was only 4GB.

Aug 28 20:19:36 kernel: Free swap  = 160kB
Aug 28 20:19:36 kernel: Total swap = 4194300kB

And here’s what earlyoom’s readme says:

earlyoom checks the amount of available memory and free swap up to 10 times a second (less often if there is a lot of free memory). By default if both are below 10%, it will kill the largest process (highest oom_score). The percentage value is configurable via command line arguments.

So the problem was that my swap file was very small so any kind of memory pressure activates earlyoom.

My Fix

I basically just made my pagefile bigger, not sure why Omarchy defaults to just 4GB on a 64GB system. Anyway here are the commands I used to do that on my system:

btrfs subvolume create /swap
chattr +C /swap # disable CoW
sudo btrfs filesystem mkswapfile -s 32G /swap/swapfile
sudo swapon -p 0 /swap/swapfile

I also learned about zram, which is cool but I just left it at the defaults for now.