[
https://issues.apache.org/jira/browse/SLING-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Bertrand Delacretaz resolved SLING-4899.
----------------------------------------
Resolution: Fixed
Assignee: Bertrand Delacretaz
At revision 1693098 this is fixed using the Sling instance ID in the jobs
storage path, as suggested.
Job content is now under paths like
{{/var/bg/jobs/6c707c04-200b-4358-92d2-02c223f9c0ee/2015/07/28/15/59/2}}
> Background servlets: potential job storage path conflicts in clustered
> environments
> -----------------------------------------------------------------------------------
>
> Key: SLING-4899
> URL: https://issues.apache.org/jira/browse/SLING-4899
> Project: Sling
> Issue Type: Bug
> Components: Extensions
> Affects Versions: Background Servlets 1.0.0
> Reporter: Marcel Reutegger
> Assignee: Bertrand Delacretaz
>
> The background servlet creates stream nodes under a configurable path with a
> local counter. When running in a cluster, it may happen that background
> servlets on multiple cluster nodes try to create the same stream node.
> {noformat}
> java.lang.IllegalArgumentException: Stream node already exists:
> /var/bg/jobs/2015/07/17/15/51/82/stream
> at
> org.apache.sling.bgservlets.impl.storage.JobDataImpl.getOutputStream(JobDataImpl.java:74)
> at org.apache.sling.bgservlets.impl.BackgroundRequestExecutionJob.<init>
> {noformat}
> It would probably be good to include the slingId in the path prefix to
> separate streams from different cluster nodes. A similar approach is
> implemented already for sling jobs.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
