[ https://issues.apache.org/jira/browse/TIKA-836?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13177546#comment-13177546 ]
Rob Tulloh commented on TIKA-836: --------------------------------- Definitely the pdf files. Here are the feeding time breakdowns by document (feeding 1 document at a time). The first pdf took 41 seconds to be parsed and indexed. The second took over 900 seconds! {noformat} Thu Dec 29 15:48:21 2011 feedBatch in: solCol2 # docs 1 bytes 239857 Thu Dec 29 15:49:03 2011 feedBatch out: solCol2 # docs 1 bytes 239857 # err 0 # millis 41570 Thu Dec 29 15:49:03 2011 Long running batch (t= 41570 ) doc 1326391801::2 mime = application/octet-stream Thu Dec 29 15:49:03 2011 feedBatch in: solCol2 # docs 1 bytes 3866139 Thu Dec 29 16:04:15 2011 feedBatch out: solCol2 # docs 1 bytes 3866139 # err 0 # millis 912770 Thu Dec 29 16:04:15 2011 Long running batch (t= 912770 ) doc 1326391801::3 mime = application/octet-stream {noformat} > parsing really slow on some documents > ------------------------------------- > > Key: TIKA-836 > URL: https://issues.apache.org/jira/browse/TIKA-836 > Project: Tika > Issue Type: Improvement > Components: parser > Affects Versions: 1.0 > Environment: CentOS 4.x/5.x/6.x > Reporter: Rob Tulloh > > We are seeing that tika sometimes takes a very long time to parse some > content (likely PDF). For example, with the following EML file that contains > 4 documents (2 PDF, 1 MS Excel, 1 text): > {noformat} > fgrep --binary-file=text Content-Type: XXX.eml > Content-Type: multipart/mixed; > Content-Type: multipart/alternative; > Content-Type: text/plain; > Content-Type: text/html; > Content-Type: application/octet-stream; > Content-Type: application/octet-stream; > Content-Type: application/vnd.ms-excel; > du -sh XXX.eml > 6.0M XXX.eml > {noformat} > Note that it takes tika nearly 30 minutes to process this content even though > the source is only 6M in size: > {noformat} > time java -Xmx2G -jar ../../tika-app-1.0.jar -m XXX.eml >meta.out > WARN - Did not found XRef object at specified startxref position 230521 > WARN - Did not found XRef object at specified startxref position 3742379 > real 29m16.913s > user 18m17.050s > sys 0m19.465s > {noformat} > Is there any way to configure tika (in particular via solr) to process files > more quickly? -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira