[
https://issues.apache.org/jira/browse/OAK-877?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13688070#comment-13688070
]
Jukka Zitting commented on OAK-877:
-----------------------------------
Right, that should be OK. However, I still have a feeling that we're treating a
symptom instead of the cause of a problem here.
>From the second graph looks like in the 10-commit case events are being
>processed only one commit per second. A better approach would be to process
>the full pending queue instead of just its first commit whenever the observer
>is invoked (i.e. make {{ChangeProcessor.run()}} loop until the
>{{getChanges()}} call returns {{null}}). My guess would be that with such a
>change the two graphs would look much more similar.
> Generating observation events takes too long when intermediate save calls are
> involved
> --------------------------------------------------------------------------------------
>
> Key: OAK-877
> URL: https://issues.apache.org/jira/browse/OAK-877
> Project: Jackrabbit Oak
> Issue Type: Improvement
> Components: core
> Reporter: Michael Dürig
> Attachments: 1_save.png, 2.save.png
>
>
> Creating observation events is much more expensive when a transaction is
> broken down through intermediate save calls compared to only having a single
> save call.
--
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
