[
https://issues.apache.org/jira/browse/CONNECTORS-1671?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17390958#comment-17390958
]
Julien Massiera commented on CONNECTORS-1671:
---------------------------------------------
[[email protected]], it is a runtime exception that occurs on Solr side,
not on connector side, and that is a cause (child) of a SolrServerException !
Furthermore, I did not create the handleRuntimeException method, it was already
there, I am just using it in another place where it can happen, for the same
reason described in the method Javadoc
{code:java}
/** Handle a SolrServerException. * These exceptions seem to be catch-all
exceptions having to do with misconfiguration or * underlying IO exceptions, or
request parsing exceptions. * If this method doesn't throw an exception, it
means that the exception should be interpreted * as meaning that the document
or action is illegal and should not be repeated. */{code}
I have just updated the Javadoc to add "or request parsing exceptions" because
this is what happens.
{code:java}
{code}
> Solr output connector behavior on some exceptions
> -------------------------------------------------
>
> Key: CONNECTORS-1671
> URL: https://issues.apache.org/jira/browse/CONNECTORS-1671
> Project: ManifoldCF
> Issue Type: Improvement
> Components: Lucene/SOLR connector
> Affects Versions: ManifoldCF 2.19
> Reporter: Julien Massiera
> Priority: Major
> Fix For: ManifoldCF next
>
> Attachments: patch-CONNECTORS-1671.txt
>
>
> In the « handleIOException » method of the « HttpPoster » class, the unknown
> case triggers a job failure despite the exception can only concern the
> document/action itself and not a problem with a potential "Solr down" issue
> (all "Solr down" issues are handled upstream)
> Same thing in the « handleSolrServerException » method
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
