Could Ubuntu "unattended-upgrade" break a running Bareos Job?
36 views
Skip to first unread message
Meggie Hallenbach
Dec 4, 2025, 7:32:57 AMDec 4
to bareos-users
Hi,
i had a job on tape running for some time, the job ended as Failure but without filling out a full log - the log ends after bringing up an new tape, labeling it and writing on it.
I now look around about the reasons of this strange behavior and found out that Ubuntu 24 "unattended-upgrade" stopped postresql and this restarted baroes...
Is this a known issue? Should i disable "unattended-upgrade"? What to do next - the tape is in the drive and waiting... the file in spoolcache is gone, the last tape is now in status "Append" - but the job is in "Error"?
Take a look at the syslog what happend after "unattended-upgrade" started while there was still a job running...
2025-12-04T06:52:06.881321+01:00 golem systemd[1]: Starting apt-daily-upgrade.service - Daily apt upgrade and clean activities...
2025-12-04T06:52:11.133866+01:00 golem apt.systemd.daily[1688941]: /usr/bin/unattended-upgrade:567: DeprecationWarning: This process (pid=1688941) is multi-threaded, use of fork() may lead to deadlocks in the child.
2025-12-04T06:52:11.134054+01:00 golem apt.systemd.daily[1688941]: pid = os.fork()
2025-12-04T06:52:12.456964+01:00 golem systemd[1]: Stopping postg...@16-main.service - PostgreSQL Cluster 16-main...
2025-12-04T06:52:13.719440+01:00 golem systemd[1]: postg...@16-main.service: Deactivated successfully.
2025-12-04T06:52:13.719644+01:00 golem systemd[1]: Stopped postg...@16-main.service - PostgreSQL Cluster 16-main.
2025-12-04T06:52:13.719983+01:00 golem systemd[1]: postg...@16-main.service: Consumed 1h 12min 1.269s CPU time, 24.1G memory peak, 0B memory swap peak.
2025-12-04T06:52:15.857599+01:00 golem systemd[1]: Reloading requested from client PID 1690422 ('systemctl') (unit apt-daily-upgrade.service)...
2025-12-04T06:52:15.857735+01:00 golem systemd[1]: Reloading...
2025-12-04T06:52:16.221026+01:00 golem systemd[1]: Reloading finished in 362 ms.
2025-12-04T06:52:16.305622+01:00 golem systemd[1]: Starting postg...@16-main.service - PostgreSQL Cluster 16-main...
2025-12-04T06:52:19.041194+01:00 golem systemd[1]: Started postg...@16-main.service - PostgreSQL Cluster 16-main.
2025-12-04T06:52:19.573795+01:00 golem dbus-daemon[1597]: [system] Activating via systemd: service name='org.freedesktop.PackageKit' unit='packagekit.service' requested by ':1.609' (uid=0 pid=1690537 comm="/usr/bin/gdbus call --system --dest org.freedeskto" label="unconfined")
2025-12-04T06:52:19.590805+01:00 golem systemd[1]: Starting packagekit.service - PackageKit Daemon...
2025-12-04T06:52:19.596688+01:00 golem PackageKit: daemon start
2025-12-04T06:52:19.628140+01:00 golem dbus-daemon[1597]: [system] Successfully activated service 'org.freedesktop.PackageKit'
2025-12-04T06:52:19.628396+01:00 golem systemd[1]: Started packagekit.service - PackageKit Daemon.
2025-12-04T06:52:20.542133+01:00 golem systemd[1]: Stopping apache2.service - The Apache HTTP Server...
2025-12-04T06:52:20.544709+01:00 golem bareos-dir: Shutting down BAREOS service: bareos-dir ...
2025-12-04T06:52:20.545145+01:00 golem systemd[1]: Stopping bareos-director.service - Bareos Director Daemon service...
2025-12-04T06:52:20.547351+01:00 golem systemd[1]: Stopping php8.3-fpm.service - The PHP 8.3 FastCGI Process Manager...
2025-12-04T06:52:20.550712+01:00 golem systemd[1]: Stopping webmin.service - Webmin server daemon...
2025-12-04T06:52:20.568043+01:00 golem apachectl[1690581]: AH00558: apache2: Could not reliably determine the server's fully qualified domain name, using 127.0.1.1. Set the 'ServerName' directive globally to suppress this message
2025-12-04T06:52:20.571010+01:00 golem systemd[1]: php8.3-fpm.service: Deactivated successfully.
2025-12-04T06:52:20.571444+01:00 golem systemd[1]: Stopped php8.3-fpm.service - The PHP 8.3 FastCGI Process Manager.
2025-12-04T06:52:20.571570+01:00 golem systemd[1]: php8.3-fpm.service: Consumed 50.874s CPU time, 27.3M memory peak, 0B memory swap peak.
2025-12-04T06:52:20.575775+01:00 golem systemd[1]: Starting php8.3-fpm.service - The PHP 8.3 FastCGI Process Manager...
2025-12-04T06:52:20.576440+01:00 golem systemd[1]: webmin.service: Main process exited, code=exited, status=1/FAILURE
2025-12-04T06:52:20.576659+01:00 golem systemd[1]: webmin.service: Failed with result 'exit-code'.
2025-12-04T06:52:20.577168+01:00 golem systemd[1]: Stopped webmin.service - Webmin server daemon.
2025-12-04T06:52:20.577275+01:00 golem systemd[1]: webmin.service: Consumed 16h 44min 43.716s CPU time, 1.1G memory peak, 0B memory swap peak.
2025-12-04T06:52:20.580559+01:00 golem systemd[1]: Starting webmin.service - Webmin server daemon...
2025-12-04T06:52:20.608840+01:00 golem systemd[1]: apache2.service: Deactivated successfully.
2025-12-04T06:52:20.609372+01:00 golem systemd[1]: Stopped apache2.service - The Apache HTTP Server.
2025-12-04T06:52:20.609500+01:00 golem systemd[1]: apache2.service: Consumed 10min 22.110s CPU time, 96.3M memory peak, 0B memory swap peak.
2025-12-04T06:52:20.635791+01:00 golem systemd[1]: Starting apache2.service - The Apache HTTP Server...
2025-12-04T06:52:20.650484+01:00 golem systemd[1]: Started php8.3-fpm.service - The PHP 8.3 FastCGI Process Manager.
2025-12-04T06:52:20.659219+01:00 golem apachectl[1690602]: AH00558: apache2: Could not reliably determine the server's fully qualified domain name, using 127.0.1.1. Set the 'ServerName' directive globally to suppress this message
2025-12-04T06:52:20.672343+01:00 golem systemd[1]: Started apache2.service - The Apache HTTP Server.
2025-12-04T06:52:23.116345+01:00 golem systemd[1]: Started webmin.service - Webmin server daemon.
2025-12-04T06:53:50.650364+01:00 golem systemd[1]: bareos-director.service: State 'stop-sigterm' timed out. Killing.
2025-12-04T06:53:50.651063+01:00 golem systemd[1]: bareos-director.service: Killing process 2333 (bareos-dir) with signal SIGKILL.
2025-12-04T06:53:50.651203+01:00 golem systemd[1]: bareos-director.service: Failed to kill control group /system.slice/bareos-director.service, ignoring: Invalid argument
2025-12-04T06:53:50.657007+01:00 golem systemd[1]: bareos-director.service: Main process exited, code=killed, status=9/KILL
2025-12-04T06:53:50.657171+01:00 golem systemd[1]: bareos-director.service: Failed with result 'timeout'.
2025-12-04T06:53:50.657807+01:00 golem systemd[1]: Stopped bareos-director.service - Bareos Director Daemon service.
2025-12-04T06:53:50.657995+01:00 golem systemd[1]: bareos-director.service: Consumed 1h 42min 39.181s CPU time, 28.8G memory peak, 0B memory swap peak.
2025-12-04T06:53:50.672347+01:00 golem systemd[1]: Started bareos-director.service - Bareos Director Daemon service.
2025-12-04T06:53:52.746393+01:00 golem systemd[1]: apt-daily-upgrade.service: Deactivated successfully.
2025-12-04T06:53:52.746576+01:00 golem systemd[1]: Finished apt-daily-upgrade.service - Daily apt upgrade and clean activities.
2025-12-04T06:53:52.746944+01:00 golem systemd[1]: apt-daily-upgrade.service: Consumed 9.842s CPU time, 194.2M memory peak, 0B memory swap peak.
2025-12-04T06:55:01.393550+01:00 golem CRON[1690935]: (root) CMD (command -v debian-sa1 > /dev/null && debian-sa1 1 1)
2025-12-04T06:58:07.165443+01:00 golem bareos-sd[1875]: bareos-sd (0): stored/spool.cc:330-11012 Fatal append error on device "lto8-0" (/dev/tape/by-id/scsi-35000e111c23a90b5-nst): ERR=Error sending Volume info to Director.
2025-12-04T06:58:55.036485+01:00 golem PackageKit: daemon quit
2025-12-04T06:58:55.042143+01:00 golem systemd[1]: packagekit.service: Deactivated successfully.
2025-12-04T07:00:00.720074+01:00 golem systemd[1]: Starting sysstat-collect.service - system activity accounting tool...
2025-12-04T07:00:00.732339+01:00 golem systemd[1]: sysstat-collect.service: Deactivated successfully.
2025-12-04T07:00:00.732613+01:00 golem systemd[1]: Finished sysstat-collect.service - system activity accounting tool.
i had a job on tape running for some time, the job ended as Failure but without filling out a full log - the log ends after bringing up an new tape, labeling it and writing on it.
I now look around about the reasons of this strange behavior and found out that Ubuntu 24 "unattended-upgrade" stopped postresql and this restarted baroes...
Is this a known issue? Should i disable "unattended-upgrade"? What to do next - the tape is in the drive and waiting... the file in spoolcache is gone, the last tape is now in status "Append" - but the job is in "Error"?
2025-12-04T06:52:06.881321+01:00 golem systemd[1]: Starting apt-daily-upgrade.service - Daily apt upgrade and clean activities...
2025-12-04T06:52:11.133866+01:00 golem apt.systemd.daily[1688941]: /usr/bin/unattended-upgrade:567: DeprecationWarning: This process (pid=1688941) is multi-threaded, use of fork() may lead to deadlocks in the child.
2025-12-04T06:52:11.134054+01:00 golem apt.systemd.daily[1688941]: pid = os.fork()
2025-12-04T06:52:12.456964+01:00 golem systemd[1]: Stopping postg...@16-main.service - PostgreSQL Cluster 16-main...
2025-12-04T06:52:13.719440+01:00 golem systemd[1]: postg...@16-main.service: Deactivated successfully.
2025-12-04T06:52:13.719644+01:00 golem systemd[1]: Stopped postg...@16-main.service - PostgreSQL Cluster 16-main.
2025-12-04T06:52:13.719983+01:00 golem systemd[1]: postg...@16-main.service: Consumed 1h 12min 1.269s CPU time, 24.1G memory peak, 0B memory swap peak.
2025-12-04T06:52:15.857599+01:00 golem systemd[1]: Reloading requested from client PID 1690422 ('systemctl') (unit apt-daily-upgrade.service)...
2025-12-04T06:52:15.857735+01:00 golem systemd[1]: Reloading...
2025-12-04T06:52:16.221026+01:00 golem systemd[1]: Reloading finished in 362 ms.
2025-12-04T06:52:16.305622+01:00 golem systemd[1]: Starting postg...@16-main.service - PostgreSQL Cluster 16-main...
2025-12-04T06:52:19.041194+01:00 golem systemd[1]: Started postg...@16-main.service - PostgreSQL Cluster 16-main.
2025-12-04T06:52:19.573795+01:00 golem dbus-daemon[1597]: [system] Activating via systemd: service name='org.freedesktop.PackageKit' unit='packagekit.service' requested by ':1.609' (uid=0 pid=1690537 comm="/usr/bin/gdbus call --system --dest org.freedeskto" label="unconfined")
2025-12-04T06:52:19.590805+01:00 golem systemd[1]: Starting packagekit.service - PackageKit Daemon...
2025-12-04T06:52:19.596688+01:00 golem PackageKit: daemon start
2025-12-04T06:52:19.628140+01:00 golem dbus-daemon[1597]: [system] Successfully activated service 'org.freedesktop.PackageKit'
2025-12-04T06:52:19.628396+01:00 golem systemd[1]: Started packagekit.service - PackageKit Daemon.
2025-12-04T06:52:20.542133+01:00 golem systemd[1]: Stopping apache2.service - The Apache HTTP Server...
2025-12-04T06:52:20.544709+01:00 golem bareos-dir: Shutting down BAREOS service: bareos-dir ...
2025-12-04T06:52:20.545145+01:00 golem systemd[1]: Stopping bareos-director.service - Bareos Director Daemon service...
2025-12-04T06:52:20.547351+01:00 golem systemd[1]: Stopping php8.3-fpm.service - The PHP 8.3 FastCGI Process Manager...
2025-12-04T06:52:20.550712+01:00 golem systemd[1]: Stopping webmin.service - Webmin server daemon...
2025-12-04T06:52:20.568043+01:00 golem apachectl[1690581]: AH00558: apache2: Could not reliably determine the server's fully qualified domain name, using 127.0.1.1. Set the 'ServerName' directive globally to suppress this message
2025-12-04T06:52:20.571010+01:00 golem systemd[1]: php8.3-fpm.service: Deactivated successfully.
2025-12-04T06:52:20.571444+01:00 golem systemd[1]: Stopped php8.3-fpm.service - The PHP 8.3 FastCGI Process Manager.
2025-12-04T06:52:20.571570+01:00 golem systemd[1]: php8.3-fpm.service: Consumed 50.874s CPU time, 27.3M memory peak, 0B memory swap peak.
2025-12-04T06:52:20.575775+01:00 golem systemd[1]: Starting php8.3-fpm.service - The PHP 8.3 FastCGI Process Manager...
2025-12-04T06:52:20.576440+01:00 golem systemd[1]: webmin.service: Main process exited, code=exited, status=1/FAILURE
2025-12-04T06:52:20.576659+01:00 golem systemd[1]: webmin.service: Failed with result 'exit-code'.
2025-12-04T06:52:20.577168+01:00 golem systemd[1]: Stopped webmin.service - Webmin server daemon.
2025-12-04T06:52:20.577275+01:00 golem systemd[1]: webmin.service: Consumed 16h 44min 43.716s CPU time, 1.1G memory peak, 0B memory swap peak.
2025-12-04T06:52:20.580559+01:00 golem systemd[1]: Starting webmin.service - Webmin server daemon...
2025-12-04T06:52:20.608840+01:00 golem systemd[1]: apache2.service: Deactivated successfully.
2025-12-04T06:52:20.609372+01:00 golem systemd[1]: Stopped apache2.service - The Apache HTTP Server.
2025-12-04T06:52:20.609500+01:00 golem systemd[1]: apache2.service: Consumed 10min 22.110s CPU time, 96.3M memory peak, 0B memory swap peak.
2025-12-04T06:52:20.635791+01:00 golem systemd[1]: Starting apache2.service - The Apache HTTP Server...
2025-12-04T06:52:20.650484+01:00 golem systemd[1]: Started php8.3-fpm.service - The PHP 8.3 FastCGI Process Manager.
2025-12-04T06:52:20.659219+01:00 golem apachectl[1690602]: AH00558: apache2: Could not reliably determine the server's fully qualified domain name, using 127.0.1.1. Set the 'ServerName' directive globally to suppress this message
2025-12-04T06:52:20.672343+01:00 golem systemd[1]: Started apache2.service - The Apache HTTP Server.
2025-12-04T06:52:23.116345+01:00 golem systemd[1]: Started webmin.service - Webmin server daemon.
2025-12-04T06:53:50.650364+01:00 golem systemd[1]: bareos-director.service: State 'stop-sigterm' timed out. Killing.
2025-12-04T06:53:50.651063+01:00 golem systemd[1]: bareos-director.service: Killing process 2333 (bareos-dir) with signal SIGKILL.
2025-12-04T06:53:50.651203+01:00 golem systemd[1]: bareos-director.service: Failed to kill control group /system.slice/bareos-director.service, ignoring: Invalid argument
2025-12-04T06:53:50.657007+01:00 golem systemd[1]: bareos-director.service: Main process exited, code=killed, status=9/KILL
2025-12-04T06:53:50.657171+01:00 golem systemd[1]: bareos-director.service: Failed with result 'timeout'.
2025-12-04T06:53:50.657807+01:00 golem systemd[1]: Stopped bareos-director.service - Bareos Director Daemon service.
2025-12-04T06:53:50.657995+01:00 golem systemd[1]: bareos-director.service: Consumed 1h 42min 39.181s CPU time, 28.8G memory peak, 0B memory swap peak.
2025-12-04T06:53:50.672347+01:00 golem systemd[1]: Started bareos-director.service - Bareos Director Daemon service.
2025-12-04T06:53:52.746393+01:00 golem systemd[1]: apt-daily-upgrade.service: Deactivated successfully.
2025-12-04T06:53:52.746576+01:00 golem systemd[1]: Finished apt-daily-upgrade.service - Daily apt upgrade and clean activities.
2025-12-04T06:53:52.746944+01:00 golem systemd[1]: apt-daily-upgrade.service: Consumed 9.842s CPU time, 194.2M memory peak, 0B memory swap peak.
2025-12-04T06:55:01.393550+01:00 golem CRON[1690935]: (root) CMD (command -v debian-sa1 > /dev/null && debian-sa1 1 1)
2025-12-04T06:58:07.165443+01:00 golem bareos-sd[1875]: bareos-sd (0): stored/spool.cc:330-11012 Fatal append error on device "lto8-0" (/dev/tape/by-id/scsi-35000e111c23a90b5-nst): ERR=Error sending Volume info to Director.
2025-12-04T06:58:55.036485+01:00 golem PackageKit: daemon quit
2025-12-04T06:58:55.042143+01:00 golem systemd[1]: packagekit.service: Deactivated successfully.
2025-12-04T07:00:00.720074+01:00 golem systemd[1]: Starting sysstat-collect.service - system activity accounting tool...
2025-12-04T07:00:00.732339+01:00 golem systemd[1]: sysstat-collect.service: Deactivated successfully.
2025-12-04T07:00:00.732613+01:00 golem systemd[1]: Finished sysstat-collect.service - system activity accounting tool.
Miguel Santos
Dec 7, 2025, 8:50:23 AM (13 days ago) Dec 7
to bareos-users
Yes, you can clearly see in the log some services are being shutdown.
[...]
2025-12-04T06:52:12.456964+01:00 golem systemd[1]: Stopping postg...@16-main.service - PostgreSQL Cluster 16-main...
2025-12-04T06:52:13.719440+01:00 golem systemd[1]: postg...@16-main.service: Deactivated successfully.
2025-12-04T06:52:13.719644+01:00 golem systemd[1]: Stopped postg...@16-main.service - PostgreSQL Cluster 16-main.
2025-12-04T06:52:13.719983+01:00 golem systemd[1]: postg...@16-main.service: Consumed 1h 12min 1.269s CPU time, 24.1G memory peak, 0B memory swap peak.
2025-12-04T06:52:15.857599+01:00 golem systemd[1]: Reloading requested from client PID 1690422 ('systemctl') (unit apt-daily-upgrade.service)...
2025-12-04T06:52:15.857735+01:00 golem systemd[1]: Reloading...
2025-12-04T06:52:16.221026+01:00 golem systemd[1]: Reloading finished in 362 ms.
2025-12-04T06:52:16.305622+01:00 golem systemd[1]: Starting postg...@16-main.service - PostgreSQL Cluster 16-main...
2025-12-04T06:52:19.041194+01:00 golem systemd[1]: Started postg...@16-main.service - PostgreSQL Cluster 16-main.
2025-12-04T06:52:13.719440+01:00 golem systemd[1]: postg...@16-main.service: Deactivated successfully.
2025-12-04T06:52:13.719644+01:00 golem systemd[1]: Stopped postg...@16-main.service - PostgreSQL Cluster 16-main.
2025-12-04T06:52:13.719983+01:00 golem systemd[1]: postg...@16-main.service: Consumed 1h 12min 1.269s CPU time, 24.1G memory peak, 0B memory swap peak.
2025-12-04T06:52:15.857599+01:00 golem systemd[1]: Reloading requested from client PID 1690422 ('systemctl') (unit apt-daily-upgrade.service)...
2025-12-04T06:52:15.857735+01:00 golem systemd[1]: Reloading...
2025-12-04T06:52:16.221026+01:00 golem systemd[1]: Reloading finished in 362 ms.
2025-12-04T06:52:16.305622+01:00 golem systemd[1]: Starting postg...@16-main.service - PostgreSQL Cluster 16-main...
2025-12-04T06:52:19.041194+01:00 golem systemd[1]: Started postg...@16-main.service - PostgreSQL Cluster 16-main.
[...]
2025-12-04T06:53:50.650364+01:00 golem systemd[1]: bareos-director.service: State 'stop-sigterm' timed out. Killing.
2025-12-04T06:53:50.651063+01:00 golem systemd[1]: bareos-director.service: Killing process 2333 (bareos-dir) with signal SIGKILL.
2025-12-04T06:53:50.651203+01:00 golem systemd[1]: bareos-director.service: Failed to kill control group /system.slice/bareos-director.service, ignoring: Invalid argument
2025-12-04T06:53:50.657007+01:00 golem systemd[1]: bareos-director.service: Main process exited, code=killed, status=9/KILL
2025-12-04T06:53:50.657171+01:00 golem systemd[1]: bareos-director.service: Failed with result 'timeout'.
2025-12-04T06:53:50.651063+01:00 golem systemd[1]: bareos-director.service: Killing process 2333 (bareos-dir) with signal SIGKILL.
2025-12-04T06:53:50.651203+01:00 golem systemd[1]: bareos-director.service: Failed to kill control group /system.slice/bareos-director.service, ignoring: Invalid argument
2025-12-04T06:53:50.657007+01:00 golem systemd[1]: bareos-director.service: Main process exited, code=killed, status=9/KILL
2025-12-04T06:53:50.657171+01:00 golem systemd[1]: bareos-director.service: Failed with result 'timeout'.
[...]
It is not an issue IMO, it is how the updates in Ubuntu work, it is better to know and plan when the updates must happen.
It is not an issue IMO, it is how the updates in Ubuntu work, it is better to know and plan when the updates must happen.
Meggie Hallenbach
Dec 7, 2025, 4:54:21 PM (12 days ago) Dec 7
to bareos-users
@ It is not an issue IMO, it is how the updates in Ubuntu work, it is better to know and plan when the updates must happen.
And as always - dont see any problem and fist of all blame the messenger... ;)
This Bareos system runs since more then 5 years on different Ubuntu LTS versions in the same configuration - i had never any problems with this "unattended-upgrade" - never! But i disabled it now - maybe something has changes in 24.4...
And to the problem:
I see that the DB-Service was restarted and the Bareos Director! But not the Storage Deamon (!).. After restarting the director the SD run into some problem and hang itself... and thats also the question - why? What happens when you restart a director with the spooled file? Is it cleaned, deleted? Does a running storage Deamon permanently need connection to the director?
And as always - dont see any problem and fist of all blame the messenger... ;)
This Bareos system runs since more then 5 years on different Ubuntu LTS versions in the same configuration - i had never any problems with this "unattended-upgrade" - never! But i disabled it now - maybe something has changes in 24.4...
And to the problem:
I see that the DB-Service was restarted and the Bareos Director! But not the Storage Deamon (!).. After restarting the director the SD run into some problem and hang itself... and thats also the question - why? What happens when you restart a director with the spooled file? Is it cleaned, deleted? Does a running storage Deamon permanently need connection to the director?
Reply all
Reply to author
Forward
0 new messages