[readme] Add notes on coredumpctl configuration #368
Description
First, thank you for your work on this. Setting it up was a breeze and it worked right away.
I have noticed absurd amounts of IO during crashes, filling the root partition to the brim. This caused drastic slowdowns due to IO pressure and filled partition. However, before I could find which files were taking up space, they always disappeared.
The solution was to limit or disable core dumping via coredumpctl configuration.
On my system, the default configuration allowed for memory dumps of a ridiculously large size to disk after an OOM kill (at least those resulting in a SIGSEGV).
After updating the configuration /etc/systemd/coredump.conf file, all issues were gone.
[Coredump]
#Storage=external
#Compress=yes
ProcessSizeMax=256M
#ExternalSizeMax=32G
#JournalSizeMax=767M
MaxUse=256M
KeepFree=2048M
#EnterNamespace=noI imagine that this is an OS-specific issue, but perhaps this can help someone in a similar situation: installing earlyoom to prevent a full memory crash, only to unknowingly find themselves in a full partition hangup.
Once again, thanks!
OS: EndeavourOS rolling rolling
Kernel: x86_64 Linux 6.18.9-zen1-2-zen
Uptime: 14h 37m
Packages: 2130
Shell: bash 5.3.9
Resolution: 3840x1080
DE: KDE
WM: KWin
GTK Theme: Breeze-Dark [GTK2], Breeze [GTK3]
Icon Theme: breeze-dark
Disk: 182G / 225G (86%)
CPU: Intel Core i7-7700HQ @ 8x 3.8GHz [61.0°C]
GPU: Mesa Intel(R) HD Graphics 630 (KBL GT2)
RAM: 7995MiB / 15841MiB