Controlling core files (Linux)
© March 2005 Tony Lawrence
Core files get created when a program misbehaves due to a bug, or a violation of the cpu or memory protection mechanisms. The operating system kills the program and creates the core file.
If you don't want core files at all, set "ulimit -c 0" in your startup files. That's the default on many systems; in /etc/profile you may find
ulimit -S -c 0 > /dev/null 2>&1
If you DO want core files, you need to reset that in your own .bash_profile:
ulimit -c 50000
would allow core files but limit them to 50,000 blocks.
You have more control of core files in /proc/sys/kernel/
For example, you can do eliminate the tagged on pid by
echo "0" > /proc/sys/kernel/core_uses_pid
Core files will then just be named "core". People do things like that so that a user can choose to put a non-writable file named "core" in directories where they don't want to generate core dumps. That could be a directory (mkdir core) or a file (touch core;chmod 000 core). I've seen it suggested that a symlink named core would redirect the dump to wherever it pointed, but I found that didn't work.
But perhaps more interesting is that you can do:
mkdir /tmp/corefiles chmod 777 /tmp/corefiles echo "/tmp/corefiles/core" > /proc/sys/kernel/core_pattern
All corefiles then get tossed to /tmp/corefiles (don't change core_uses_pid if you do this).
Test this with a simple script:
# script that dumps core kill -s SIGSEGV $$
But wait, there's more (if your kernel is new enough). From "man proc":
/proc/sys/kernel/core_pattern
This file (new in Linux 2.5) provides finer control over the
form of a core filename than the obsolete
/proc/sys/kernel/core_uses_pid file described below. The name
for a core file is controlled by defining a template in
/proc/sys/kernel/core_pattern. The template can contain %
specifiers which are substituted by the following values when
a core file is created:
%% A single % character
%p PID of dumped process
%u real UID of dumped process
%g real GID of dumped process
%s number of signal causing dump
%t time of dump (secs since 0:00h, 1 Jan 1970)
%h hostname (same as the 'nodename'
returned by uname(2))
%e executable filename
A single % at the end of the template is dropped from the core
filename, as is the combination of a % followed by any character
other than those listed above. All other characters in the
template become a literal part of the core filename. The maximum
size of the resulting core filename is 64 bytes. The default
value in this file is "core". For backward compatibility, if
/proc/sys/kernel/core_pattern does not include "%p" and
/proc/sys/kernel/core_uses_pid is non-zero, then .PID will be
appended to the core filename.
If you are running a Linux kernel that doesn't support this, you'll get no core files at all, which is also what happens if the directory in core_pattern doesn't exist or isn't writable by the user dumping core. So that's yet another way to not dump core for certain users: set core_pattern to a directory that they can't write to, and give write permission to the users who you do want to create core files.
The "ulimit" can do much more: see Understanding ulimit
Got something to add? Send me email.
(OLDER) <- More Stuff -> (NEWER) (NEWEST)
Printer Friendly Version
-> -> Controlling core files (Linux)
Inexpensive and informative Apple related e-books:
El Capitan: A Take Control Crash Course
Take Control of Preview
Take Control of Upgrading to El Capitan
iOS 10: A Take Control Crash Course
Sierra: A Take Control Crash Course
More Articles by Tony Lawrence © 2015-11-22 Tony Lawrence
What do such machines really do? They increase the number of things we can do without thinking. Things we do without thinking — there's the real danger. (Frank Herbert)
This post tagged:
Skills Tests
Unix/Linux Book Reviews
My Unix/Linux Troubleshooting Book
This site runs on Linode
Sat Feb 9 06:57:20 2008: 3628 anonymous
This topic was really helpful to know about core files in linux/unix. Appreciated!!
Mon Jun 23 16:11:03 2008: 4363 anonymous
Ultimate guide. It does take some time to figure out why there are no core files on ubuntu. apport is not set correctly by default to handle core files in server installations. This write up is really getting to the bottom of it.
Mon Dec 8 22:18:25 2008: 4884 anonymous
I found this page via Google, and just wanted to say that I found it really useful - no need for further research. First click, and satisfied - awesome! thanks!
Mon Feb 9 05:46:55 2009: 5350 ZhichangYu
Really good guide! However It's a pity that it doesn't mention that /etc/sysctl.conf controls all settings under /proc/sys.
Mon Feb 9 12:12:45 2009: 5351 TonyLawrence
A pity?
:-)
I guess sysctl and sysctl.conf are fairly standard now but weren't when this was written. Not all distros used sysctl.conf then and I wouldn't bet my life that all do today.
I think it is probably true that sysctl.conf is a fairly safe assumption now, but more generally you should never assume that because you know something about one distro at one point in time that it applies to other distros or even that same distro at another point in time.
If you pawed through the thousands of posts here, you'd find lots of examples like that: things that were once true that aren't now, things that mention /proc but not sysctl, things that mention sysctl but not sysctl.conf, things that have changed, things that now don't work the way they did when the post was written - it's impossible for me to keep up with and that's why I REALLY APPRECIATE COMMENTS LIKE THIS.
Even when laced with pity :-)
Fri Apr 17 19:13:29 2009: 6216 Slinky
Use this to find core files and remove them:
find . | egrep "\/core\.[0-9]+$" | xargs rm -f
This works well as it finds only core files.
Courtesy of (link)
Fri Apr 17 19:42:12 2009: 6217 TonyLawrence
Well, sure..
But that's not what this article was about. :-)
Thu Jun 4 07:10:10 2009: 6437 DomenPuncer
From manual: "Values are in 1024-byte increments..."
so ulimit -c 50000 is about 50 MB
Thu Jun 4 10:08:55 2009: 6438 TonyLawrence
Ooops - thanks!
Fri Jun 19 04:44:19 2009: 6516 anonymous
This article looks informative.
Could you please tell me, how can we store core dumps in NVRAM ?
Fri Jun 19 09:26:49 2009: 6517 TonyLawrence
I can't imagine why you want to do that, but obviously you'd need th NVRAM mounted and you'd use the "/proc/sys/kernel/core_pattern" explained above.
Given the usual size of NVRAM and the overhead of filesystems, you couldn't store much..
Tue Dec 8 17:29:20 2009: 7725 anonymous
I am using Kill -9 to teminate a pid and not getting a core file. Are core files not generated for kill -9 signals? (it does for other Kill signals)
Tue Dec 8 18:33:24 2009: 7726 TonyLawrence
"man signal" will tell you which signals dump core.
Mon Mar 8 08:41:01 2010: 8192 RickvanderZwet
Very nice and informative explanation. Works great under RHEL4 for example
Wed Dec 15 14:58:09 2010: 9168 vmguy
To complete the article, the following points should be made:
- To print the call stacks from a core file using gdb:
thread apply all bt
- If any of the function names in the core file are annotated as "??",
it is an indication that the core file may have been truncated on
the filesystem. Another ulimit value controls this:
ulimit -f unlimited
- The core file may also be truncated because there is insufficient space
on the filesystem to write it ... especially if this is a 64-bit app.
cat /proc/sys/kernel/core_pattern
Example: if this contains /work, then check the filesystem for space
df -k /work
Freespace is displayed as a percentage of total space and as the
number of free 1K blocks. Do the math. If the calculation shows
1.5GB is available, then a 2GB core file won't fit there.
Nice article, Tony.
Wed Dec 15 15:06:40 2010: 9169 TonyLawrence
Thanks for the additional notes!
Mon Jan 31 22:11:04 2011: 9265 anonymous
Your article was very informative. However, there seems to be an issue with the core_uses_pid variable. In the distro I am using (RHEL5), this has no effect. In searching the net, I found reference to this some time ago, along with mention of a patch to fix it, but this does not seem to have taken place. I have core_uses_pid set to 0, and no %p in core_pattern, but I still get the PID appended to the core file name.
Mon Jan 31 22:31:55 2011: 9266 TonyLawrence
I'm not surprised. Nobody every accused Linux kernel developers of being consistent.
Sat Feb 5 01:58:20 2011: 9278 Venkat
Awesome aticle. Our Test server was crashing everyday but there was no core file being generated. This article helped him overwrite those settings. THANKS.
Sun Nov 6 14:25:28 2011: 10127 Neha
HI
Actually a lot of core files are being generated on my database server on path /var/opt/OV/tmp after every 15 min.How to let it not be generated
Sun Nov 6 14:33:58 2011: 10128 TonyLawrence
Umm, that's what this whole page is about?
------------------------
Printer Friendly Version
Controlling core files (Linux) Copyright © March 2005 Tony Lawrence
Have you tried Searching this site?
This is a Unix/Linux resource website. It contains technical articles about Unix, Linux and general computing related subjects, opinion, news, help files, how-to's, tutorials and more.
Contact us
Printer Friendly Version