[
https://issues.apache.org/jira/browse/SLING-2493?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Alexander Klimetschek updated SLING-2493:
-----------------------------------------
Description:
MapEntries [0] does first a scan in doInit() and then registers the event
listener for sling:vanityPath nodes. Now if a observation event comes in
between that, it gets lost - the old content was read in the initial scan and
is cached.
It looks like we had such a case while both bundle and content were updated "at
the same time". Possibly the registerService() method could be blocked for a
while if many other services are restarted in such a update scenario involving
multiple bundles.
I think the observation listener for such cases should always be registered
*before* the initial scan.
[0]
http://svn.apache.org/repos/asf/sling/trunk/bundles/jcr/resource/src/main/java/org/apache/sling/jcr/resource/internal/helper/MapEntries.java
was:
MapEntries [0] does first a scan in doInit() and then starts a thread for
getting observation changes for sling:vanityPath nodes. Now if a observation
event comes in between that, it gets lost - the old content was read in the
initial scan and is cached. It looks like we had such a case while both bundle
and content were updated "at the same time".
I think the observation listener for such cases should always be registered
*before* the initial scan.
[0]
http://svn.apache.org/repos/asf/sling/trunk/bundles/jcr/resource/src/main/java/org/apache/sling/jcr/resource/internal/helper/MapEntries.java
> sling:vanityPaths: observation listener should be started before initial scan
> on bundle activation to avoid missing changes
> ---------------------------------------------------------------------------------------------------------------------------
>
> Key: SLING-2493
> URL: https://issues.apache.org/jira/browse/SLING-2493
> Project: Sling
> Issue Type: Bug
> Components: JCR
> Affects Versions: JCR Resource 2.1.0
> Reporter: Alexander Klimetschek
> Priority: Minor
>
> MapEntries [0] does first a scan in doInit() and then registers the event
> listener for sling:vanityPath nodes. Now if a observation event comes in
> between that, it gets lost - the old content was read in the initial scan and
> is cached.
> It looks like we had such a case while both bundle and content were updated
> "at the same time". Possibly the registerService() method could be blocked
> for a while if many other services are restarted in such a update scenario
> involving multiple bundles.
> I think the observation listener for such cases should always be registered
> *before* the initial scan.
> [0]
> http://svn.apache.org/repos/asf/sling/trunk/bundles/jcr/resource/src/main/java/org/apache/sling/jcr/resource/internal/helper/MapEntries.java
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
