LINUX.IE, website of the Irish Linux Users' Group
Tux rules!

   
Home
New Users
Articles
Download
Projects
Community
Vendors

  Print Version
Email to...
 
Archives:


planetILUG

Recent News

News Archive


Join the
ILUG
on FaceBook


Join the
ILUG
on LinkedIn


Join the
ILUG SETI
Group



















 
 :: Mailing Lists

[ILUG] kernel upgrade and swap

[ILUG] kernel upgrade and swap

Paul Jakma paul at jakma.org
Fri Apr 6 01:01:28 IST 2001


On Thu, 5 Apr 2001, John P. Looney wrote:

>  Just had my first experience with it yesterday. Wine went nuts, and
> grabbed all ram. The machine started getting slow...then Linux killed
> xchat, xmms and gnome-terminal on me, to free up memory for Wine. Nice
> one!

or you could have set reasonable per-process limits, on RH ->
/etc/security/limits.conf, eg:

@users          hard    core    100000
@users          soft    rss     99000
*               hard    rss     110000
@users          soft    nproc   100
@users          hard    nproc   200
@users          soft    nofile  1000
@users          hard    nofile  3000
@users          soft    data    80000
@users          hard    data    110000
#@users         soft    stack   16384
#@users         hard    stack   32768

and make sure that your xdm/kdm/gdm pam config files include
pam_limits.so under session. On RH7 most services reference a common
pam config file, system-auth.. make sure it has:

session     required      /lib/security/pam_limits.so

and hey presto, as long as your limits are sane you will probably
never hit the linux OOM killer.

> Kate

To answer Paul Collins:

> I wish he'd fix the OOM killer and get the balancing right first.

balancing isn't the problem.

> I'm getting seriously tired of watching him and the MM "team"
> futz this.

futz what, implementation of code that is guaranteed to never please
everyone all the time?

> His main response to people complaining/reporting bugs (he seems
> to equate the two) is "show me the code", even when it's obvious
> bugs in code *he* wrote.

<rant>

well to be fair, how the fsck are you ever going to write a perfect
process killer? think about it... it's impossible unless you have (to
borrow a phrase from mr. lyda) esp.c.

what annoys me the most is how people whine about this on
linux-kernel, when there exists a mechanism in linux to help avoid
hitting OOM (and the attendent kernel mm/oom_kill.c) -> per process
limits - and they don't bother setting it up.

So netscape locked up your box, or the kernel killed xmms, xchat,
etc.. instead of the rampant wine.. well if you had a reasonable
limit set then wine would have hit it and died before your box became
unreasonably slow and your other processes would be fine.

now, per-process limits aren't perfect - they won't stop a user or
process deliberately trying to hit OOM - but they do make a huge
difference in everday usage - use them.

</rant>

regards,
-- 
Paul Jakma	paul at clubi.ie	paul at jakma.org
PGP5 key: http://www.clubi.ie/jakma/publickey.txt
-------------------------------------------
Fortune:
I cannot draw a cart, nor eat dried oats; If it be man's work I will do it.






More information about the ILUG mailing list
Read this without the formatting.
                                                                                                    

 

Hosted by HEAnet


Maintained by the ILUG website team. The aim of Linux.ie is to support and help commercial and private users of Linux in Ireland. You can display ILUG news in your own webpages, read backend information to find out how. Networking services kindly provided by HEAnet, server kindly donated by Dell. Linux is a trademark of Linus Torvalds, used with permission. No penguins were harmed in the production or maintenance of this highly praised website. Looking for the Indian Linux Users' Group? Try here. If you've read all this and aren't a lawyer: you should be!
RSS Version
Powered by Dell