Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.debian.user > #209834 > unrolled thread
| Started by | Martin T <m4rtntns@gmail.com> |
|---|---|
| First post | 2019-06-12 11:50 +0200 |
| Last post | 2019-06-12 16:40 +0200 |
| Articles | 4 — 2 participants |
Back to article view | Back to linux.debian.user
How to disable automatic core dumps in Debian 9? Martin T <m4rtntns@gmail.com> - 2019-06-12 11:50 +0200
Re: How to disable automatic core dumps in Debian 9? Reco <recoverym4n@enotuniq.net> - 2019-06-12 14:00 +0200
Re: How to disable automatic core dumps in Debian 9? Martin T <m4rtntns@gmail.com> - 2019-06-12 16:20 +0200
Re: How to disable automatic core dumps in Debian 9? Reco <recoverym4n@enotuniq.net> - 2019-06-12 16:40 +0200
| From | Martin T <m4rtntns@gmail.com> |
|---|---|
| Date | 2019-06-12 11:50 +0200 |
| Subject | How to disable automatic core dumps in Debian 9? |
| Message-ID | <y883g-5Is-5@gated-at.bofh.it> |
Hi, how to disable all automatic core dumps upon the process termination in Debian 9? As Debian does not seem to use systemd-coredump, then I set the "kernel.core_pattern=|/bin/false" kernel configuration in sysctl.conf and following pam_limits.so module configuration: $ grep -w core /etc/security/limits.conf # - core - limits the core file size (KB) * hard core 0 root hard core 0 #root hard core 100000 $ When I execute a program, which crashes, then while the "core dumped" is printed: $ ./testprogram Floating point exception (core dumped) $ ..then no actual core dump does not seem to be created. Is the configuration above enough to disable core dumps or am I missing something? thanks, Martin
[toc] | [next] | [standalone]
| From | Reco <recoverym4n@enotuniq.net> |
|---|---|
| Date | 2019-06-12 14:00 +0200 |
| Message-ID | <y8a53-6St-1@gated-at.bofh.it> |
| In reply to | #209834 |
Hi. On Wed, Jun 12, 2019 at 12:49:14PM +0300, Martin T wrote: > As Debian does not seem to use systemd-coredump, It's a feature, not a flaw. > then I set the "kernel.core_pattern=|/bin/false" kernel configuration > in sysctl.conf sysctl -w kernel.core_pattern=/dev/null There's no need to pipe the core to /bin/false if you don't need the core in the first place. > and following pam_limits.so module configuration: > > $ grep -w core /etc/security/limits.conf > # - core - limits the core file size (KB) > * hard core 0 > root hard core 0 > #root hard core 100000 > $ rgrep core /etc/security/limits* Would be more correct. Files at /etc/security/limits.d/ can override limits.conf. > Ween I execute a program, which crashes, then while the "core dumped" > is printed: A simple 'ulimit -c' would suffice. Reco
[toc] | [prev] | [next] | [standalone]
| From | Martin T <m4rtntns@gmail.com> |
|---|---|
| Date | 2019-06-12 16:20 +0200 |
| Message-ID | <y8cgx-8lB-7@gated-at.bofh.it> |
| In reply to | #209840 |
Reco, thanks for reply! > rgrep core /etc/security/limits* > Would be more correct. Files at /etc/security/limits.d/ can override limits.conf. Indeed. Output of "rgrep core /etc/security/limits*" is following: $ rgrep core /etc/security/limits* /etc/security/limits.conf:# - core - limits the core file size (KB) /etc/security/limits.conf:* hard core 0 /etc/security/limits.conf:root hard core 0 /etc/security/limits.conf:#root hard core 100000 $ So in short, the configuration in my initial e-mail was enough to disable core dumps? After reading the "man 5 core", I also set the "kernel.core_uses_pid" to 0. thanks, Martin
[toc] | [prev] | [next] | [standalone]
| From | Reco <recoverym4n@enotuniq.net> |
|---|---|
| Date | 2019-06-12 16:40 +0200 |
| Message-ID | <y8czT-8sa-5@gated-at.bofh.it> |
| In reply to | #209842 |
Hi. On Wed, Jun 12, 2019 at 05:14:09PM +0300, Martin T wrote: > Reco, > > thanks for reply! > > > rgrep core /etc/security/limits* > > Would be more correct. Files at /etc/security/limits.d/ can override limits.conf. > > Indeed. Output of "rgrep core /etc/security/limits*" is following: > > $ rgrep core /etc/security/limits* > /etc/security/limits.conf:# - core - limits the core file size (KB) > /etc/security/limits.conf:* hard core 0 > /etc/security/limits.conf:root hard core 0 > /etc/security/limits.conf:#root hard core 100000 > $ > > So in short, the configuration in my initial e-mail was enough to > disable core dumps? I prefer to do it this way, but that's a matter of a personal preference (note that it works for root too): * - core 0 Overriding kernel.core_pattern is way too limiting for me. I need those core files in certain cases (but live happily without them by default). > After reading the "man 5 core", I also set the "kernel.core_uses_pid" to 0. Zero is a default value for this in Debian anyway. So - can't hurt, but I doubt it'll change something for you. Reco
[toc] | [prev] | [standalone]
Back to top | Article view | linux.debian.user
csiph-web