Services not restarting on reboot

[Koldnitz]

Hello,

I have been messing with sensei this afternoon and in doing so have rebooted my router about 10 times.

I have noticed that netdata and syslog-ng sometimes do not come up with the server.  They have a few times, but mostly they do not.

This never happened previously on 20.1, I rebooted the router a whole bunch on that series because I hosed my configuration a few times.

Has anyone noticed this? 

I have not changed my config since 20.1 other than to turn off circular logging.

I am hoping nothing is corrupted because I really do not want to have to rebuild this thing again.

Cheers,

[Koldnitz]

Is this happening to anyone other than me and jerry89?

Jerry89 posed in the postfix not logging thread and seems to have disabled circular logging as well.

Should I not disable circular logging?

I read in multiple places it was on it way out and was going to be removed.

I am not sure whats going on with netdata doing a complete reinstall to see if that fixes it

Cheers,

[FullyBorked]

I'm having similar issues, syslog-ng never starts on it's own.  I always have to manually start it. 

[Koldnitz]

I just reset logs and that didn't fix it.

I turned circular logging back on and while syslogd comes up immediately, remote syslog-ng does not, and must be manually started still.

I guess maybe this is just a bug (I hope something on my system didn't corrupt)?

Cheers,

[FullyBorked]

I wouldn't worry about corruption.  I just installed new hardware with a fresh install of 20.7 and am experiencing the same behavior.  My guess is it's a bug, you could probably use monit to monitor the PID and start it if it's not running.  Probably what I'll do. 

[FullyBorked]

If I can figure out the syntax to start the service I'll have monit up and working,  so far the commands I can find don't work.  I can monitor the service and test for it's existence, but can't make it it start. 

Code Select Expand

failed to start (exit status 0) -- '/usr/local/sbin/configctl syslog-ng start': Action not found

[FullyBorked]

This is what I have so far.  I cannot figure out the start syntax to put in monit that will actually work.  Maybe someone else can help me make it farther.

Service Test Settings:

Name: Does not exist
Condition: does not exist
Action: Start

Service Settings:

Name: Check_SyslogNG_Process
Type: Process
PID File: /var/run/syslog-ng.pid
Start: ?
Stop: ?
Tests: Does not exist

[hushcoden]

Same issue here, syslog-ng never starts on it's own and I've to manually start it...

[baqwas]

I had to reboot 20.7 twice and in both occasions, as others have reported, I had to manually start syslog-ng. At least I can start this service unlike snmpd which refuses to start at all.

Kind regards.

[aimdev]

I had an issue like this after the upgrade, so I re-installed and then setup with a backup.
Both syslogs are running, not checked if they were the issue behind a massive use of the disk over time.

https://forum.opnsense.org/index.php?topic=18358.0