You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
class { 'icingadb::redis':
manage_repos => false,
bind => ['127.0.0.1', '::1'],
port => 6380,
}
Upgrade from 7.0.15-1+debian12 => 7.2.6-2+debian12
What are you seeing
Icingadb-redis won't start with the following error:
... systemd[1]: Starting icingadb-redis.service - Redis database for Icinga DB...
... icingadb-redis-server[3754696]: *** FATAL CONFIG FILE ERROR (Redis 7.2.6) ***
... icingadb-redis-server[3754696]: Can't open the log file: Read-only file system
... systemd[1]: icingadb-redis.service: Main process exited, code=exited, status=1/FAILURE
... systemd[1]: icingadb-redis.service: Failed with result 'exit-code'.
... systemd[1]: Failed to start icingadb-redis.service - Redis database for Icinga DB.
What behaviour did you expect instead
icingadb-redis to start successfully after an upgrade.
Output log
See above.
Any additional information you'd like to impart
According to this forum post, the latest icingdb-redis package took out ReadWritePaths=-/var/log/icingadb-redis from the service unit file shipped with the package and set logfile="" in the default configuration file.
This module uses puppet-redis, which in turn expects a log file to be set - I don't think there's a way to override this. This module also sets manage_service_file => false, so the service unit as shipped is used.
We've worked around this by adding a drop-in file via our profile class:
systemd::dropin_file { 'redis-logfile.conf':
unit => 'icingadb-redis.service',
content => "[Service]\nReadWritePaths=-/var/log/icingadb-redis\n",
}
I wasn't sure what the best place for this was, but I thought I'd open an issue here for visibility. There may be a way of working around this in a different fashion with the modules as-is, but this is what we did.
The text was updated successfully, but these errors were encountered:
Hi, I haven't any problem with these versions if I use the default logdir /var(log/icingadb-redis. The unit file of icingadb-redis-7.2.6-2 has the setting ReadWritePaths=-/var/log/icingadb-redis.
The initial release, 7.2.6-1, had omitted the line, it looks like it was restored in 7.2.6-2. This got crossed over in my testing, so you're quite right, this is no longer an issue.
Yes. You're right. In the previous version of the package I noticed this issue I did my best to get it fixed by icinga.
To log to journald I opened a PR: voxpupuli/puppet-redis#534
Affected Puppet, Ruby, OS and module versions/distributions
How to reproduce (e.g Puppet code you use)
Upgrade from 7.0.15-1+debian12 => 7.2.6-2+debian12
What are you seeing
Icingadb-redis won't start with the following error:
What behaviour did you expect instead
icingadb-redis to start successfully after an upgrade.
Output log
See above.
Any additional information you'd like to impart
According to this forum post, the latest icingdb-redis package took out
ReadWritePaths=-/var/log/icingadb-redis
from the service unit file shipped with the package and setlogfile=""
in the default configuration file.This module uses puppet-redis, which in turn expects a log file to be set - I don't think there's a way to override this. This module also sets
manage_service_file => false
, so the service unit as shipped is used.We've worked around this by adding a drop-in file via our profile class:
I wasn't sure what the best place for this was, but I thought I'd open an issue here for visibility. There may be a way of working around this in a different fashion with the modules as-is, but this is what we did.
The text was updated successfully, but these errors were encountered: