collectd is unable to start

[AniTexs]

I have issues with starting collectd in order to get my Cloudron server online again.

Im starting a new support thread since I was unable to find another reliable support thread to use.

Here is the Journalctl output

root@my:/home/yellowtent/platformdata/nginx# journalctl -u collectd
-- Logs begin at Wed 2022-06-15 05:35:14 UTC, end at Wed 2022-06-15 19:17:32 UTC. --
Jun 15 05:35:14 URL collectd[71032]: ERROR: dlopen("/usr/lib/collectd/python.so") failed: /usr/lib/collectd/python.so: undefined symbol: PyFloat_Type. The >
Jun 15 05:35:14 URL collectd[71032]: Error: Parsing the config file failed!
Jun 15 05:35:14 URL systemd[1]: collectd.service: Main process exited, code=exited, status=1/FAILURE
Jun 15 05:35:14 URL systemd[1]: collectd.service: Failed with result 'exit-code'.
Jun 15 05:35:14 URL systemd[1]: Failed to start Statistics collection and monitoring daemon.

[AniTexs]

@girish It works after inserting the LD_PRELOAD

root@my:~# /home/yellowtent/box/setup/start.sh
2022-06-16T14:58:38 ==> start: Cloudron Start
media:x:500:
2022-06-16T14:58:38 ==> start: Configuring docker
Synchronizing state of apparmor.service with SysV service script with /lib/systemd/systemd-sysv-install.
Executing: /lib/systemd/systemd-sysv-install enable apparmor
2022-06-16T14:58:40 ==> start: Ensuring directories
2022-06-16T14:58:40 ==> start: Configuring journald
2022-06-16T14:58:41 ==> start: Setting up unbound
2022-06-16T14:58:41 ==> start: Adding systemd services
Synchronizing state of unbound.service with SysV service script with /lib/systemd/systemd-sysv-install.
Executing: /lib/systemd/systemd-sysv-install enable unbound
Synchronizing state of cron.service with SysV service script with /lib/systemd/systemd-sysv-install.
Executing: /lib/systemd/systemd-sysv-install enable cron
2022-06-16T14:58:48 ==> start: Configuring sudoers
2022-06-16T14:58:48 ==> start: Configuring collectd
Job for collectd.service failed because the control process exited with error code.
See "systemctl status collectd.service" and "journalctl -xe" for details.
root@my:~# systemctl restart collectd
Job for collectd.service failed because the control process exited with error code.
See "systemctl status collectd.service" and "journalctl -xe" for details.
root@my:~#

The content of /etc/default/collectd

  GNU nano 4.8                                                            /etc/default/collectd                                                                      # /etc/default/collectd

# 0: start collectd on boot, 1: do not start collectd on boot
# default: 0
DISABLE=0

# 0: start collectd in stand-alone mode, 1: monitor collectd using collectdmon
# default: 1
USE_COLLECTDMON=1

# number of seconds to wait for collectd to shut down
# default: 30
MAXWAIT=30

# 0: do not enable core-files, 1: enable core-files ... if collectd crashes
# default: 0
ENABLE_COREFILES=0

After inserting the line

# /etc/default/collectd

# 0: start collectd on boot, 1: do not start collectd on boot
# default: 0
DISABLE=0

# 0: start collectd in stand-alone mode, 1: monitor collectd using collectdmon
# default: 1
USE_COLLECTDMON=1

# number of seconds to wait for collectd to shut down
# default: 30
MAXWAIT=30

# 0: do not enable core-files, 1: enable core-files ... if collectd crashes
# default: 0
ENABLE_COREFILES=0

LD_PRELOAD=/usr/lib/python3.8/config-3.8-x86_64-linux-gnu/libpython3.8.so

root@my:~# nano /etc/default/collectd
root@my:~# systemctl restart collectd
root@my:~#

[girish]

@AniTexs which ubuntu version are you using? Did you upgrade ubuntu?

[AniTexs]

@girish Yes. from 18.04 to 20.04, followed the guide, everything was working after the upgrade, this happend 2 days after the upgrade

(Have to wait 360 seconds to reply....)

[girish]

@AniTexs I think maybe the collectd config got overwritten during the upgrade. If you run /home/yellowtent/box/setup/start.sh , this will put the expected collectd config.

Also, check the content of /etc/default/collectd . It should contain:

LD_PRELOAD=/usr/lib/python3.8/config-3.8-x86_64-linux-gnu/libpython3.8.so

After that, systemctl restart collectd.

[AniTexs]

@girish It works after inserting the LD_PRELOAD

root@my:~# /home/yellowtent/box/setup/start.sh
2022-06-16T14:58:38 ==> start: Cloudron Start
media:x:500:
2022-06-16T14:58:38 ==> start: Configuring docker
Synchronizing state of apparmor.service with SysV service script with /lib/systemd/systemd-sysv-install.
Executing: /lib/systemd/systemd-sysv-install enable apparmor
2022-06-16T14:58:40 ==> start: Ensuring directories
2022-06-16T14:58:40 ==> start: Configuring journald
2022-06-16T14:58:41 ==> start: Setting up unbound
2022-06-16T14:58:41 ==> start: Adding systemd services
Synchronizing state of unbound.service with SysV service script with /lib/systemd/systemd-sysv-install.
Executing: /lib/systemd/systemd-sysv-install enable unbound
Synchronizing state of cron.service with SysV service script with /lib/systemd/systemd-sysv-install.
Executing: /lib/systemd/systemd-sysv-install enable cron
2022-06-16T14:58:48 ==> start: Configuring sudoers
2022-06-16T14:58:48 ==> start: Configuring collectd
Job for collectd.service failed because the control process exited with error code.
See "systemctl status collectd.service" and "journalctl -xe" for details.
root@my:~# systemctl restart collectd
Job for collectd.service failed because the control process exited with error code.
See "systemctl status collectd.service" and "journalctl -xe" for details.
root@my:~#

The content of /etc/default/collectd

  GNU nano 4.8                                                            /etc/default/collectd                                                                      # /etc/default/collectd

# 0: start collectd on boot, 1: do not start collectd on boot
# default: 0
DISABLE=0

# 0: start collectd in stand-alone mode, 1: monitor collectd using collectdmon
# default: 1
USE_COLLECTDMON=1

# number of seconds to wait for collectd to shut down
# default: 30
MAXWAIT=30

# 0: do not enable core-files, 1: enable core-files ... if collectd crashes
# default: 0
ENABLE_COREFILES=0

After inserting the line

# /etc/default/collectd

# 0: start collectd on boot, 1: do not start collectd on boot
# default: 0
DISABLE=0

# 0: start collectd in stand-alone mode, 1: monitor collectd using collectdmon
# default: 1
USE_COLLECTDMON=1

# number of seconds to wait for collectd to shut down
# default: 30
MAXWAIT=30

# 0: do not enable core-files, 1: enable core-files ... if collectd crashes
# default: 0
ENABLE_COREFILES=0

LD_PRELOAD=/usr/lib/python3.8/config-3.8-x86_64-linux-gnu/libpython3.8.so

root@my:~# nano /etc/default/collectd
root@my:~# systemctl restart collectd
root@my:~#

[necrevistonnezr]

@girish said in collectd is unable to start:

@AniTexs I think maybe the collectd config got overwritten during the upgrade. If you run /home/yellowtent/box/setup/start.sh , this will put the expected collectd config.

Also, check the content of /etc/default/collectd . It should contain:

LD_PRELOAD=/usr/lib/python3.8/config-3.8-x86_64-linux-gnu/libpython3.8.so

After that, systemctl restart collectd.

Similar issue after upgrading from Ubuntu 20.04 to 22.04 - but despite the reference to libpython3.8.so being present in /etc/default/collectd , it seems that this file is no longer present...

However, what is present is /usr/lib/python3.10/config-3.10-x86_64-linux-gnu/libpython3.10.so - after referencing this file in /etc/default/collectd (i.e. adding LD_PRELOAD=/usr/lib/python3.10/config-3.10-x86_64-linux-gnu/libpython3.10.so), and restarting the collectd service, it starts running again.

... but will this change persist?

[girish]

@necrevistonnezr said in collectd is unable to start:

... but will this change persist?

yes, that's the correct fix. I had this line in the ubuntu 20 upgrade guide, but it's missing in the ubuntu 22 guide. I have added it now.

[mathieuepitech]

I have the same issue, and i tried to add the LD_PRELOAD think, but i got a parsing error in the file.
Does anyone have already saw that ?

May 10 15:14:23 sanchezm systemd[1]: Starting Statistics collection and monitoring daemon...
May 10 15:14:23 sanchezm collectd[6484]: [2023-05-10 15:14:23] plugin_load: plugin "logfile" successfully loaded.
May 10 15:14:23 sanchezm collectd[6484]: Error: Parsing the config file failed!
May 10 15:14:23 sanchezm systemd[1]: collectd.service: Main process exited, code=exited, status=1/FAILURE
May 10 15:14:23 sanchezm systemd[1]: collectd.service: Failed with result 'exit-code'.
May 10 15:14:23 sanchezm systemd[1]: Failed to start Statistics collection and monitoring daemon.
May 10 15:14:25 sanchezm systemd[1]: Stopped Statistics collection and monitoring daemon.

After some research, i found that the issue come from the Import "du" in the collectd.conf file.
Does anyone know how to fix that ?

<Plugin python>
    # https://blog.dbrgn.ch/2017/3/10/write-a-collectd-python-plugin/
    ModulePath "/home/yellowtent/box/setup/start/collectd/"
    LogTraces false # enable this to get traces in /var/log/collectd.log
    Interactive false

    Import "df"

#    Import "du"
    <Module du>
        <Path>
            Instance maildata
            Dir "/home/yellowtent/boxdata/mail"
        </Path>
        <Path>
            Instance boxdata
            Dir "/home/yellowtent/boxdata"
            Exclude "mail"
        </Path>
        <Path>
            Instance platformdata
            Dir "/home/yellowtent/platformdata"
        </Path>
    </Module>
</Plugin>

[girish]

@mathieuepitech Interesting. du is a custom plugin we wrote. Can you tell which ubuntu version you are on? Also, the python version (3.8 or 3.10?).

[mathieuepitech]

@girish I just upgrade to ubuntu 20.04.6 LTS
And python 3.8.
I just check back the collectd.conf, and I don"t have anymore the module du.
I upgrade to cloudron 7.4.1.

[girish]

@mathieuepitech Ah, I didn't realize you were on a older Cloudron version. Indeed, du.py is now gone.