Skip to content
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

[Documentation for newcomers / new users of earlyoom] #211

Open
rubyFeedback opened this issue Jul 2, 2020 · 7 comments
Open

[Documentation for newcomers / new users of earlyoom] #211

rubyFeedback opened this issue Jul 2, 2020 · 7 comments

Comments

@rubyFeedback
Copy link

Hey earlyoom devs,

Recently Fedora announced a (some?) plan to move/use/integrate earlyoom.

I am an average user really (mostly), using a bit of ruby and other languages,
but I am not an expert when it comes to kernel-related stuff. I think many
other users are no experts either.

I like to tinker with stuff though (one reason why I don't use systemd; tinkering
without it is much easier, and I try to keep the complexity very low at all times
so I understand what is going on). This indirectly brings me to earlyoom.

I read through the README here (for the most part) and a lot of it sounds
familiar to me. I remember compiling LLVM and the old kernel I use right
now (slackware; but I know how to update, will do so in the next days)
also killed firefox and palemoon when all 4 CPUs were working hard.

This was unexpected to me. It also was annoying since I (a) had to
restart the browser, but more importantly (b) found it annoying that
the kernel can decide what to kill. For example, I'd be more happy
to say something like "compile with only one CPU, slowly, without
killing my system as I work on it right now". And I think that such a
use case may fit to what earlyoom is trying.

Which brings me to this feature request. :)

Would it be possible for you guys to perhaps, on some weekend
in the future, go through the main README and improve on it
in particular for newcomers who are not totally dumb but not
very clever or knowledeable in regards to the kernel-internals?

My hope here is that slightly better documentation may help me,
and others, to use earlyoom effectively. This is actually more or
less the first time I heard about it; I'll compile earlyoom soon
and have a look, but every bit of high quality documentation
would help me.

At any rate if this is too much work please just feel free to
close this issue request at any moment in time - it is more
important that a project can retain momentum, and motivated
devs, I understand that very well.

@hakavlad
Copy link
Contributor

hakavlad commented Jul 2, 2020

Hi! Answers to what questions would you like to receive?

@antofthy
Copy link

Greetings... I found earlyoom running on my latest fedora (v33) and wondered what it was.
So I looked at the man page...

It starts with the phrase...
The oom-killer generally has a bad reputation among Linux users.

This was completely unhelpfull to anyone who has no idea what the program is, does, or what it is for!
I mean at no point in the manpage does it even say What is a OOM!
And while after having to read the whole manpage (which I should not have needed to do) I now have a fair idea what the program does, I still don't know what "OOM" actually means or stands for!

I eventually decided to take a guess that the URL given in BUGs points to the software source, though the URL is not properly given in the manpage. However the main README file was just a copy of the manpage, and still does not tell me anything new.

I believe this is what @rubyFeedback is getting at. Documentation is not new user friendly and assumes you already have an understanding of the program and problems it is trying to solve. Thus the issue.

@jayenashar
Copy link

Hi @antofthy OOM = Out Of Memory, and the oom-killer is a linux thing that kills processes when the system is out of memory. My issue with the oom-killer is that it runs too late, about an hour after my system becomes unresponsive, so earlyoom kills processes before the system is out of memory.

If you find this information useful, do you mind to make a pull request and update the man page and/or README?

@arch-user-france1
Copy link

Hi @antofthy OOM = Out Of Memory, and the oom-killer is a linux thing that kills processes when the system is out of memory. My issue with the oom-killer is that it runs too late, about an hour after my system becomes unresponsive, so earlyoom kills processes before the system is out of memory.

If you find this information useful, do you mind to make a pull request and update the man page and/or README?

problem isn't it runs too late the problem is that it runs when already out of memory not when running out of memory so I recommend all kernel developers to make a room-killer and f* of that not-well-working oom-killer. Earlyoom can solve this kinda

@hakavlad
Copy link
Contributor

I recommend all kernel developers to make a room-killer and f* of that not-well-working oom-killer

Use this patch if you want to fix the kernel.

@arch-user-france1
Copy link

Well that's not a fix it is an improovment
But many kernels like xanmod2 should have it
Am I correct?

@arch-user-france1
Copy link

Hey earlyoom devs,

Recently Fedora announced a (some?) plan to move/use/integrate earlyoom.

I am an average user really (mostly), using a bit of ruby and other languages,
but I am not an expert when it comes to kernel-related stuff. I think many
other users are no experts either.

I like to tinker with stuff though (one reason why I don't use systemd; tinkering
without it is much easier, and I try to keep the complexity very low at all times
so I understand what is going on). This indirectly brings me to earlyoom.

I read through the README here (for the most part) and a lot of it sounds
familiar to me. I remember compiling LLVM and the old kernel I use right
now (slackware; but I know how to update, will do so in the next days)
also killed firefox and palemoon when all 4 CPUs were working hard.

This was unexpected to me. It also was annoying since I (a) had to
restart the browser, but more importantly (b) found it annoying that
the kernel can decide what to kill. For example, I'd be more happy
to say something like "compile with only one CPU, slowly, without
killing my system as I work on it right now". And I think that such a
use case may fit to what earlyoom is trying.

Which brings me to this feature request. :)

Would it be possible for you guys to perhaps, on some weekend
in the future, go through the main README and improve on it
in particular for newcomers who are not totally dumb but not
very clever or knowledeable in regards to the kernel-internals?

My hope here is that slightly better documentation may help me,
and others, to use earlyoom effectively. This is actually more or
less the first time I heard about it; I'll compile earlyoom soon
and have a look, but every bit of high quality documentation
would help me.

At any rate if this is too much work please just feel free to
close this issue request at any moment in time - it is more
important that a project can retain momentum, and motivated
devs, I understand that very well.

I think it is a plan not some plan. Some would be plural - but I am not sure if some would be correct

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

5 participants