I have a dozen of identical services configured in monit.
When I enable alerts for check process on these services and alert condition arises in all of the services I receive a dozen of identical alerts.
Is there a way to do some kind of aggregation (based on process-check group probably) so that I receive only a single alert within one monit cycle?
When service is described with both ensure => running and subscribed to config file changes if both conditions fulfil (i.e. it is not running and config file changed) it gets (re)started twice. Is there a way to make it start only once?
To clarify, I have the following type:
service { "puppet":
ensure => running,
enable => true,
hasrestart => true,
subscribe => File["/etc/puppet/puppet.conf"],
} # service
When both /etc/puppet/puppet.conf is changed and service puppet is not running I see the following in puppet output (with --debug specified):
....
debug: /Stage[main]/Puppet/Service[puppet]/subscribe: subscribes to File[/etc/puppet/puppet.conf]
notice: /Stage[main]/Puppet/File[/etc/puppet/puppet.conf]/content: content changed '{md5}c39350dd1e135cf9fffd14ff167d6655' to '{md5}845a4aac5e83ab3f8cd9efd80ce17772'
....
info: /etc/puppet/puppet.conf: Scheduling refresh of Service[puppet]
debug: /etc/puppet/puppet.conf: The container Class[Puppet] will propagate my refresh event
info: /etc/puppet/puppet.conf: Scheduling refresh of Service[puppet]
debug: Service[puppet](provider=debian): Executing '/etc/init.d/puppet status'
debug: Service[puppet](provider=debian): Executing '/etc/init.d/puppet start'
notice: /Stage[main]/Puppet/Service[puppet]/ensure: ensure changed 'stopped' to 'running'
debug: /Stage[main]/Puppet/Service[puppet]: The container Class[Puppet] will propagate my refresh event
debug: Service[puppet](provider=debian): Executing '/etc/init.d/puppet status'
debug: Service[puppet](provider=debian): Executing '/etc/init.d/puppet restart'
err: /Stage[main]/Puppet/Service[puppet]: Failed to call refresh: Could not restart Service[puppet]: Execution of '/etc/init.d/puppet restart' returned 1: at /etc/puppet/modules/puppet/manifests/init.pp:32
This obviously fails because of the following:
root:~# /etc/init.d/puppet restart ; /etc/init.d/puppet restart
* Restarting puppet agent [ OK ]
* Restarting puppet agent [fail]
Is there a way to cope with that?
I have got a HDD drive which has gone into pre-failure state with big count of relocated sectors.
However I figured out that the problem is possibly with a HDD slot (cable) but not with a disk — when I plugged other HDDs to this slot they started working slowly and showing some seek errors and pending sectors. But these errors eventually disappear if I plug these drives to another controller. So I think that the pre-fail HDD here is not really failing.
Is it possible to force SMART to re-check Relocated and Offline Uncorrectable sectors to recover them if the failure has gone?
What is the SCSI RAID channels? I.e. what is the difference between, say, 1-channel RAID controller and 2-channel RAID controller?
Is it for data throughput, number of RAID arrays I can create on controller, number of disks I can plug into controller or what else?