[
https://issues.apache.org/jira/browse/SLING-4407?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14322804#comment-14322804
]
Carsten Ziegeler commented on SLING-4407:
-----------------------------------------
[~egli] Nice idea, right now we create OSGi events for
created/start/finished/failed - assigned is identical with created due to the
push model we're using; reassigned is missing.
We could create OSGi events for that as well and could then implement an audit
log based on the events.
Or we could really do the logging at the low level.
Now, I'm wondering what the best approach wrt clustered installations is - as
the action can take place on any node in a cluster, we could log on each node
whenever an action is taken; of course this would require log aggregation later
on to get the full picture. Or we could log on the master only, but then the
master might change. WDYT?
> introduce an audit-style logger to job handling
> -----------------------------------------------
>
> Key: SLING-4407
> URL: https://issues.apache.org/jira/browse/SLING-4407
> Project: Sling
> Issue Type: Improvement
> Components: Extensions
> Affects Versions: Event 3.5.0
> Reporter: Stefan Egli
>
> To help with debugging issues around the job handling we could introduce an
> 'audit-style logger' which - when enabled - would log the most important
> actions available on each job, ie:
> * when the job is created
> * when it is assigned, reassigned
> * when it is finished/fails
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
