-
Notifications
You must be signed in to change notification settings - Fork 159
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
del -k #80
Comments
Hmm, why? |
It is dangerous option and works worse then userspace OOMK. |
I wonder if anybody uses it. I usually don't want to break existing users. But then, the only way to find out is to remove it and see if we get complaints. |
sudo earlyoom -k
As you see, -k invokes unexpected behavior and output spam. It even worst then SIGTERM. |
and more if start quick memhog
|
I start |
then I stopped oom-trigger. |
At least unpredictable behavior should be mentioned in the documentation, if you do not want to delete the option. |
IMHO -k users will not lose anything if they start using the userspace OOM preventer instead of -k. |
Please try echo 1 > /proc/sys/kernel/sysrq it is possible that calling the kernel oom killer is blocked by default |
Hmm this only documented in the source code, https://github.com/rfjakob/earlyoom/blob/master/kill.c Ok, i guess nobody uses -k. |
The only possible meaning of this option is the demonstration of inefficiency of native OOMK.
Nothing happened. |
Magic SysRq enabled by boot cmdline on my Debian. |
Just out of curiosity, what does dmesg say when earlyoom tries to call the kernel oom killer? |
|
Intel integrated graphics? Looks the same at my box: https://gist.github.com/rfjakob/346b7dc611fc3cdf4011 |
Yes. |
The kernel oom-killer has many problems, has no features, and is probably used by nobody at all. Quoting from the (remove) function comment: * Invoke the kernel oom killer by writing "f" into /proc/sysrq-trigger * * This approach has a few problems: * 1) It is disallowed by default (even for root) on Fedora 20. * You have to first write "1" into /proc/sys/kernel/sysrq to enable the "f" * trigger. * 2) The Chrome web browser assigns a penalty of 300 onto its own tab renderer * processes. On an 8GB RAM machine, this means 2400MB, and will lead to every * tab being killed before the actual memory hog * See https://code.google.com/p/chromium/issues/detail?id=333617 for more info * 3) It is broken in 4.0.5 - see * f7e2ced #80
Removed in 368c172 . |
IMO -k option is deprecated and should be deleted at all.
The text was updated successfully, but these errors were encountered: