Bug#954255: incron: Buildd fails for fixed bug #930526 - incron creates zombie processes
Hi Emmanuel, * Peter Collinson [2020-03-19 11:32+] Package: incron Version: 0.5.12-1 Severity: normal Dear Maintainer, I was investigating why incron-0.5.12-1 causes loads of 'zombie' processes, and found the bug report #930526. The necessary fix and upgrade to 0.5.12-2 is in 'unstable' but has not migrated to testing, presumably because it doesn't compile on systems that don't support the Linux specific file notify feature. See: https://buildd.debian.org/status/package.php?p=incron=sid This makes the package hard to obtain for mortals using amd64 systems. Is there a way where the package can declare that it will not compile on certain systems? If so can this be applied, so the revised package can migrate into the open, and preferably become part of a buster update? The current version of incron is unsafe to use unless you reschedule restarts periodically to make the dead processes go away and seriously impacts on the use of the package. incron 0.5.12-2 still hasn't made it to debian buster, and incron continues to be unusable on debian stable systems :/ Is there nothing that can be done to make this version available to debian buster? Thanks -- Gerald Oskoboiny http://www.w3.org/People/Gerald/
Bug#954255: incron: Buildd fails for fixed bug #930526 - incron creates zombie processes
Package: incron Version: 0.5.12-1 Severity: normal Dear Maintainer, I was investigating why incron-0.5.12-1 causes loads of 'zombie' processes, and found the bug report #930526. The necessary fix and upgrade to 0.5.12-2 is in 'unstable' but has not migrated to testing, presumably because it doesn't compile on systems that don't support the Linux specific file notify feature. See: https://buildd.debian.org/status/package.php?p=incron=sid This makes the package hard to obtain for mortals using amd64 systems. Is there a way where the package can declare that it will not compile on certain systems? If so can this be applied, so the revised package can migrate into the open, and preferably become part of a buster update? The current version of incron is unsafe to use unless you reschedule restarts periodically to make the dead processes go away and seriously impacts on the use of the package. Thanks -- System Information: Debian Release: 10.3 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Kernel: Linux 4.19.0-8-amd64 (SMP w/4 CPU cores) Kernel taint flags: TAINT_WARN Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8), LANGUAGE=en_GB:en (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages incron depends on: ii adduser 3.118 ii init-system-helpers 1.56+nmu1 ii libc62.28-10 ii libgcc1 1:8.3.0-6 ii libstdc++6 8.3.0-6 ii lsb-base 10.2019051400 incron recommends no packages. incron suggests no packages. -- Configuration Files: /etc/incron.allow [Errno 13] Permission denied: '/etc/incron.allow' /etc/incron.deny [Errno 13] Permission denied: '/etc/incron.deny' -- no debconf information
Bug#930526: incron creates zombie processes
Hi Michael,
On Thu, Nov 28, 2019 at 12:22:19PM +0100, Michael Prokop wrote:
[...]
> Is there any ETA regarding upload towards unstable and stable from
> your side?
First, I apologize for the late reply.
I've just uploaded incron in unstable with your patch (thanks).
I've also started to prepare an uploaded for the next buster point
release. I will keep you informed.
(As you asked me I've also removed you from the Uploaders)
Regards,
--
Emmanuel Bouthenot
mail: kolter@{openics,debian}.orggpg: 4096R/0x929D42C3
xmpp: kol...@im.openics.org irc: kolter@{freenode,oftc}
Bug#930526: incron creates zombie processes
* Michael Prokop [Wed Nov 20, 2019 at 09:58:03AM +0100]: > * Michael Prokop [Wed Nov 13, 2019 at 08:34:48AM +0100]: > > I created a merge request for Emmanuel: > > https://salsa.debian.org/kolter/incron/merge_requests/1 > > Emmanuel, please consider merging this and upload the package > > towards unstable and stable ASAP (incron is really unuseable as-is). > > Let me know if you need anything else or if anything should prevent > > you from taking care. > Ping, Emmanuel? Thanks for merging the MR, Emmanuel. Is there any ETA regarding upload towards unstable and stable from your side? regards -mika- signature.asc Description: Digital signature
Bug#930526: incron creates zombie processes
Hi, * Michael Prokop [Wed Nov 13, 2019 at 08:34:48AM +0100]: > I created a merge request for Emmanuel: > https://salsa.debian.org/kolter/incron/merge_requests/1 > Emmanuel, please consider merging this and upload the package > towards unstable and stable ASAP (incron is really unuseable as-is). > Let me know if you need anything else or if anything should prevent > you from taking care. Ping, Emmanuel? regards -mika- signature.asc Description: Digital signature
Bug#930526: incron creates zombie processes
Hi, (Cc-ing original bug reporter as well who triggered this, thanks!) * Alexander Brüning [Sun Oct 27, 2019 at 08:41:29PM +]: > This is still an issue on 0.5.12-1 > These commits apparently fix it > https://github.com/ar-/incron/commit/f45c2f5ac4baea99b48e99a713d1f4ec1854aa76 ACK, though only https://github.com/ar-/incron/pull/42/commits/196975d26fd04176a1c877fa3c404efd8103c9c2 is relevant in the version present in Debian (the other commit is already present). A customer of mine was affected as well by this and I created a custom local package for buster which includes the fix, works fine and as expected there. Though it definitely makes sense to a) upload this to unstable and also b) provide a package upload to stable (to be part of the upcoming point release of buster) to fix this. I created a merge request for Emmanuel: https://salsa.debian.org/kolter/incron/merge_requests/1 Emmanuel, please consider merging this and upload the package towards unstable and stable ASAP (incron is really unuseable as-is). Let me know if you need anything else or if anything should prevent you from taking care. regards -mika- signature.asc Description: Digital signature
Bug#930526: incron creates zombie processes
This is still an issue on 0.5.12-1 These commits apparently fix it https://github.com/ar-/incron/commit/f45c2f5ac4baea99b48e99a713d1f4ec1854aa76 A thread on Stackoverflow suggests exiting with a non-zero exit code from your shell scripts. https://stackoverflow.com/questions/53722957/incrond-processes-with-shell-script-only-exit-if-script-exit-code-is-1
Bug#930526: incron creates zombie processes
* Dmitry Nezhevenko [2019-06-14 16:53+0300] Package: incron Version: 0.5.12-1 After updating two machines from current stable to buster I've got a lot of zombie processes caused by incron. We have this issue as well after updating our hosts to buster. It makes incron unusable on our systems -- after a few days there are thousands of zombie processes. :( I am not including more details because the issue seems to be well-known, but please let me know if more info would be useful. Thanks, -- Gerald Oskoboiny http://www.w3.org/People/Gerald/
Bug#930526: incron creates zombie processes
Dear Maintainer, we did not have much luck with the patch provided, however we do experience the same issue. This upstream commit from an unmerged PR helped though: https://github.com/ar-/incron/pull/56/commits/0a8cb83983394c2bd36105fcae5c56f89e931686 In its current state this package seriously compromises system stability. On Fri, 14 Jun 2019 16:53:07 +0300 Dmitry Nezhevenko wrote: > Package: incron > Version: 0.5.12-1 > Severity: important > Tags: patch > > Dear Maintainer, > > After updating two machines from current stable to buster I've got a lot > of zombie processes caused by incron. > > It looks like this known issue and was already fixed: > https://bugzilla.redhat.com/show_bug.cgi?id=1656939 > > I'm attaching copy of patch from Fedora package that fixes this issue for > me. > > It would be cool to have this applied to Buster release > > -- System Information: > Debian Release: 10.0 > APT prefers unstable-debug > APT policy: (500, 'unstable-debug'), (500, 'unstable'), (1, 'experimental-debug'), (1, 'experimental') > Architecture: amd64 (x86_64) > Foreign Architectures: i386 > > Kernel: Linux 5.1.9 (SMP w/8 CPU cores) > Kernel taint flags: TAINT_WARN, TAINT_OOT_MODULE > Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en_US:en (charmap=UTF-8) > Shell: /bin/sh linked to /bin/dash > Init: systemd (via /run/systemd/system) > > Versions of packages incron depends on: > ii adduser 3.118 > ii init-system-helpers 1.56+nmu1 > ii libc6 2.28-10 > ii libgcc1 1:8.3.0-7 > ii libstdc++6 8.3.0-7 > ii lsb-base 10.2019051400 > > incron recommends no packages. > > incron suggests no packages. > > -- Configuration Files: > /etc/incron.allow [Errno 13] Permission denied: '/etc/incron.allow' > /etc/incron.deny [Errno 13] Permission denied: '/etc/incron.deny' > > -- no debconf information > > -- > WBR, Dmitry -- Kind regards, Kim-Alexander Brodowski IServ GmbH Entwicklung Bültenweg 73 38106 Braunschweig GERMANY Phone: +49 531 22 43 666-0 Mobile: +49 152 55 17 55 16 Fax: +49 531 22 43 666-9 E-Mail:kim.brodow...@iserv.eu Internet: https://iserv.eu USt-IdNo. DE265149425 | Amtsgericht Braunschweig | HRB 201822 Managing directors: Benjamin Heindl, Martin Hüppe, Jörg Ludwig Privacy: https://iserv.eu/privacy
Bug#930526: incron creates zombie processes
I'm also seeing this after a buster upgrade. Applying this patch and rebuilding the package seems to have done the trick -- Met vriendelijke groet / Regards, Herman van Rink Initfour websolutions
Bug#930526: incron creates zombie processes
Package: incron
Version: 0.5.12-1
Severity: important
Tags: patch
Dear Maintainer,
After updating two machines from current stable to buster I've got a lot
of zombie processes caused by incron.
It looks like this known issue and was already fixed:
https://bugzilla.redhat.com/show_bug.cgi?id=1656939
I'm attaching copy of patch from Fedora package that fixes this issue for
me.
It would be cool to have this applied to Buster release
-- System Information:
Debian Release: 10.0
APT prefers unstable-debug
APT policy: (500, 'unstable-debug'), (500, 'unstable'), (1,
'experimental-debug'), (1, 'experimental')
Architecture: amd64 (x86_64)
Foreign Architectures: i386
Kernel: Linux 5.1.9 (SMP w/8 CPU cores)
Kernel taint flags: TAINT_WARN, TAINT_OOT_MODULE
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8),
LANGUAGE=en_US:en (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)
Versions of packages incron depends on:
ii adduser 3.118
ii init-system-helpers 1.56+nmu1
ii libc62.28-10
ii libgcc1 1:8.3.0-7
ii libstdc++6 8.3.0-7
ii lsb-base 10.2019051400
incron recommends no packages.
incron suggests no packages.
-- Configuration Files:
/etc/incron.allow [Errno 13] Permission denied: '/etc/incron.allow'
/etc/incron.deny [Errno 13] Permission denied: '/etc/incron.deny'
-- no debconf information
--
WBR, Dmitry
diff -Nur incron-0.5.12.orig/icd-main.cpp incron-0.5.12/icd-main.cpp
--- incron-0.5.12.orig/icd-main.cpp 2019-01-05 11:43:19.722640603 -0800
+++ incron-0.5.12/icd-main.cpp 2019-01-05 11:45:41.236340779 -0800
@@ -105,6 +105,7 @@
g_fFinish = true;
break;
case SIGCHLD:
+ do {} while (waitpid((pid_t)-1, 0, WNOHANG) > 0); /* Prevent zombies */
// first empty pipe (to prevent internal buffer overflow)
do {} while (read(g_cldPipe[0], g_cldPipeBuf, CHILD_PIPE_BUF_LEN) > 0);
