[
https://issues.apache.org/jira/browse/SOLR-310?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mike Klaas updated SOLR-310:
----------------------------
Attachment: maxPendingDeletes.patch
q&d patch that implements the above. Called it maxPendingDeletes, since the
complexity/memory being bounded is determined by that, not pending docs.
> Deleting pending handling can cause OOM
> ---------------------------------------
>
> Key: SOLR-310
> URL: https://issues.apache.org/jira/browse/SOLR-310
> Project: Solr
> Issue Type: Bug
> Reporter: Mike Klaas
> Assignee: Mike Klaas
> Fix For: 1.3
>
> Attachments: maxPendingDeletes.patch
>
>
> > Mike:
> >> Could happen when doDeleting the pending docs too. James: try
> >> sending commit every 500k docs or so.
> Yonik:
> > Hmmm, right... some of the memory usage will be related to the treemap
> > keeping track of deleted items, and some of it will be related to the
> > reader itself (the term index, and the norms).
> > Perhaps we should have some sort of limit on the number of deletions
> > we cache so people don't have to worry about that part.
> Good idea. Perhaps a solrconfig.xml setting, <maxPendingDocs>? Default to
> 100k?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
