[
https://issues.apache.org/jira/browse/CONNECTORS-1610?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16849479#comment-16849479
]
Karl Wright commented on CONNECTORS-1610:
-----------------------------------------
Hello Olivier,
ManifoldCF does not retry documents "endlessly". Eventually it will give up
and either abort the job or skip the document.
We cannot fix Solr problems by forcing ManifoldCF to ignore all Solr errors.
That would be what we'd have to do if we accepted 500 as being "ignore the
document". I would suggest opening a Solr ticket for this particular problem.
> handle error 500 in WindowsShare repository connector
> ------------------------------------------------------
>
> Key: CONNECTORS-1610
> URL: https://issues.apache.org/jira/browse/CONNECTORS-1610
> Project: ManifoldCF
> Issue Type: Bug
> Reporter: Olivier Tavard
> Priority: Major
>
> Hi,
>
> I have a question regarding error 500 in the WindowsShare repository
> connector.
>
> I recently noticed that I have a problem with a particular file that
> contains metadata with non ASCII characters. My pipeline in MCF basically
> contains the embedded Tika and the data is sent to Solr.
>
> For this particular file (it is a autocad file btw) there is an error 500
> that occurs in Solr. This happens after the embedded Tika in MCF has
> extracted content+metadata and has sent it to Solr.
>
> The job does not stop and the file is sent many times to Solr which responds
> with the same error again and again :
> The detail of the error in Solr is :
> null:org.apache.commons.fileupload.FileUploadException: Header section has
> more than 10240 bytes (maybe it is not properly terminated)
> at
> org.apache.commons.fileupload.FileUploadBase.parseRequest(FileUploadBase.java:362)
> at
> org.apache.commons.fileupload.servlet.ServletFileUpload.parseRequest(ServletFileUpload.java:115)
>
>
> In the MCF simple history, I can see that the same file is retried endlessly
> (see below) and the job is still running.
> Is there a chance to change this behavior to skip the file in this case or at
> least to stop the job after a certain number of retries ?
>
> PS : I sent 2 times an email in the dev mailing list but the emails never
> showed up, it is why I have created directly this issue.
>
> Thanks,
>
> Olivier
>
>
> {code:java}
> 27/05/19 14:24:48 document ingest (DatafariSolrNoTika)
> file://///x.x.x.x/testfiler0...
> .dwg
> 500 34 369 Error from server at http://127.0.0.1:8983/solr/FileShare:
> Expected mime type application/octet-stream but got application/json. {
> "error":{ "msg":"Header section has more than 10240 bytes (maybe it is not
> properly terminated)",
> "trace":"org.apache.commons.fileupload.FileUploadException: Header section
> has more than 10240 bytes (maybe it is not properly terminated)\n\tat
> org.apache.commons.fileupload.FileUploadBase.parseRequest(FileUploadBase.java:362)\n\tat
>
> org.apache.commons.fileupload.servlet.ServletFileUpload.parseRequest(ServletFileUpload.java:115)\n\tat
>
> org.apache.solr.servlet.SolrRequestParsers$MultipartRequestParser.parseParamsAndFillStreams(SolrRequestParsers.java:602)\n\tat
>
> org.apache.solr.servlet.SolrRequestParsers$StandardRequestParser.parseParamsAndFillStreams(SolrRequestParsers.java:784)\n\tat
> org.apache.solr.servlet.So 27/05/19 14:24:47 extract [Tika]
> file://///x.x.x.x/testfiler0...
> .dwg
> OK 34 74
> 27/05/19 14:23:45 document ingest (DatafariSolrNoTika)
> file://///x.x.x.x/testfiler0...
> .dwg
> 500 34 393 Error from server at http://127.0.0.1:8983/solr/FileShare:
> Expected mime type application/octet-stream but got application/json. {
> "error":{ "msg":"Header section has more than 10240 bytes (maybe it is not
> properly terminated)",
> "trace":"org.apache.commons.fileupload.FileUploadException: Header section
> has more than 10240 bytes (maybe it is not properly terminated)\n\tat
> org.apache.commons.fileupload.FileUploadBase.parseRequest(FileUploadBase.java:362)\n\tat
>
> org.apache.commons.fileupload.servlet.ServletFileUpload.parseRequest(ServletFileUpload.java:115)\n\tat
>
> org.apache.solr.servlet.SolrRequestParsers$MultipartRequestParser.parseParamsAndFillStreams(SolrRequestParsers.java:602)\n\tat
> org.apache.solr.servlet{code}
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
