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
Assignee: gbeutner
Status: Resolved (closed on 2016-05-11 14:10:04 +00:00)
Target Version: 2.4.8
Last Update: 2016-05-11 14:10:04 +00:00 (in Redmine)
Backport?: Not yet backported
Include in Changelog: 0
Hi,
A coworker of mine spent some time finding a way to determine if an icinga 2 slave / agent instance refused reloading because of configuration issues and his most practical solution was checking for the corresponding logentry.
I can imagine it would be fairly easy to implement an internal variable which indicates if the current configuration (on the filesystem on the master or the current synchronised version on other hosts) is valid and icinga 2 can reload. Whenever the log message about an aborted restart due to invalid config is written to the logfile this variable could be set to false. When a reload / restart succeeds the variable is set to true. It imagine it would be fairly easy to create an embedded check like icinga or cluster that checks the state of this variable.
This way we still don't know if a slave / agent received the current configuration from the config master but at least we could detect problems with instances not running the synced configuration due to errors in it
This issue has been migrated from Redmine: https://dev.icinga.com/issues/11129
Created by twidhalm on 2016-02-10 13:04:20 +00:00
Assignee: gbeutner
Status: Resolved (closed on 2016-05-11 14:10:04 +00:00)
Target Version: 2.4.8
Last Update: 2016-05-11 14:10:04 +00:00 (in Redmine)
Hi,
A coworker of mine spent some time finding a way to determine if an icinga 2 slave / agent instance refused reloading because of configuration issues and his most practical solution was checking for the corresponding logentry.
I can imagine it would be fairly easy to implement an internal variable which indicates if the current configuration (on the filesystem on the master or the current synchronised version on other hosts) is valid and icinga 2 can reload. Whenever the log message about an aborted restart due to invalid config is written to the logfile this variable could be set to false. When a reload / restart succeeds the variable is set to true. It imagine it would be fairly easy to create an embedded check like icinga or cluster that checks the state of this variable.
This way we still don't know if a slave / agent received the current configuration from the config master but at least we could detect problems with instances not running the synced configuration due to errors in it
Cheers,
Thomas
Changesets
2016-05-11 14:07:28 +00:00 by gbeutner 1ad4d9c
2016-05-12 09:11:03 +00:00 by gbeutner ff24863
Relations:
The text was updated successfully, but these errors were encountered: