[
https://issues.apache.org/jira/browse/SOLR-1939?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13836065#comment-13836065
]
Shalin Shekhar Mangar commented on SOLR-1939:
---------------------------------------------
Thanks for the patch John and sorry that I'm so late in reviewing it.
I think a better way to handle this would be to throw a SEVERE level wrapped
exception from the SolrWriter itself instead of handling it in DocBuilder. That
is also how data source level exceptions are handled.
The problem with this patch is that it aborts on any IOException which means
that a connect or read timeout can also abort indexing. So we need to be very
careful about throwing this only if Solr runs out of disk space.
> DataImportHandler reports success after running out of disk space
> -----------------------------------------------------------------
>
> Key: SOLR-1939
> URL: https://issues.apache.org/jira/browse/SOLR-1939
> Project: Solr
> Issue Type: Bug
> Components: contrib - DataImportHandler
> Affects Versions: 1.4
> Reporter: Wojtek Piaseczny
> Labels: newdev
> Attachments: SOLR-1939.patch, notes_for_SOLR1939.rtf
>
>
> DataImportHandler reports success after running out of disk space
> (.properties file is updated with new time stamps and status page reports
> success). Out of disk space exceptions should be treated more like
> datasource-level exceptions than like document-level exceptions (i.e. they
> should cause the import to report failure).
> Original discussion here:
> http://lucene.472066.n3.nabble.com/DataImportHandler-and-running-out-of-disk-space-td835125.html
--
This message was sent by Atlassian JIRA
(v6.1#6144)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
