[
https://issues.apache.org/jira/browse/JCR-3427?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Marcel Reutegger updated JCR-3427:
----------------------------------
Attachment: JCR-3427.patch
Proposed changes.
With those changes, the check for more records is moved outside the read lock
and gives competing threads a chance to acquire it.
> JCR-3138 may cause resource starvation
> --------------------------------------
>
> Key: JCR-3427
> URL: https://issues.apache.org/jira/browse/JCR-3427
> Project: Jackrabbit Content Repository
> Issue Type: Bug
> Components: clustering, jackrabbit-core
> Affects Versions: 2.3.2, 2.3.3, 2.3.4, 2.3.5, 2.3.6, 2.3.7
> Reporter: Marcel Reutegger
> Priority: Minor
> Attachments: JCR-3427.patch
>
>
> The new feature introduced with JCR-3138 may cause resource starvation when
> enabled. Consider a two node cluster where one of them writes continuously.
> The second cluster node will start synchronizing those changes from the
> journal while it holds a read lock on the journal. This lock is never
> released when new records are continuously available. Write operations on
> second cluster node will therefore be blocked forever because it cannot get
> hold of the journal lock.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
