Hi Tibor, Thanks to Ferran Jorba aid, I think we have located the problem (but I haven't solved it). We found that there are a record (record 468) had a pdf file uploaded without extension. Because that, Cds-invenio give us problems in the fulltext indexation when it try to index the ARTICLE-2010-019_1 file I deleted the record 468 but there where other records related in other tables of the database. Ferran teached me some sql instructions to delete the record in all tables, and now the instruction '$ echo "delete from bibdoc where id=684" | dbexec' have no results. But, I have still de same problem and Bibindex try still to index this deleted record!!!! I send you the end of bibsched_task_1905.log, perhaps you can identify some error we didn't found 2010-02-05 10:30:25 --> Estimated runtime: 0.5 minutes 2010-02-05 10:30:25 --> idxWORD09F for 467-467 is in consistent state 2010-02-05 10:30:25 --> idxWORD09F adding records #467-#467 started 2010-02-05 10:30:29 --> idxWORD09F adding records #467-#467 ended 2010-02-05 10:30:29 --> idxWORD09F normal wordtable flush started 2010-02-05 10:30:29 --> ...updating 3923 words into idxWORD09F started 2010-02-05 10:30:29 --> ......processed 392/3923 words 2010-02-05 10:30:29 --> ......processed 784/3923 words 2010-02-05 10:30:30 --> ......processed 1176/3923 words 2010-02-05 10:30:30 --> ......processed 1568/3923 words 2010-02-05 10:30:30 --> ......processed 1960/3923 words 2010-02-05 10:30:30 --> ......processed 2352/3923 words 2010-02-05 10:30:30 --> ......processed 2744/3923 words 2010-02-05 10:30:31 --> ......processed 3136/3923 words 2010-02-05 10:30:31 --> ......processed 3528/3923 words 2010-02-05 10:30:31 --> ......processed 3920/3923 words 2010-02-05 10:30:31 --> ...updating 3923 words into idxWORD09F ended 2010-02-05 10:30:31 --> ...updating reverse table idxWORD09R started 2010-02-05 10:30:31 --> ...updating reverse table idxWORD09R ended 2010-02-05 10:30:31 --> idxWORD09F normal wordtable flush ended 2010-02-05 10:30:31 --> idxWORD09F backing up 2010-02-05 10:30:31 --> 400 records took 159.2 seconds to complete.(151 recs/min) 2010-02-05 10:30:31 --> Estimated runtime: 0.5 minutes 2010-02-05 10:30:31 --> idxWORD09F for 468-468 is in consistent state 2010-02-05 10:30:31 --> idxWORD09F adding records #468-#468 started 2010-02-05 10:30:31 --> idxWORD09F normal wordtable flush started 2010-02-05 10:30:31 --> ...updating 0 words into idxWORD09F started 2010-02-05 10:30:31 --> ...updating 0 words into idxWORD09F ended 2010-02-05 10:30:31 --> ...updating reverse table idxWORD09R started 2010-02-05 10:30:31 --> ...updating reverse table idxWORD09R ended 2010-02-05 10:30:31 --> idxWORD09F normal wordtable flush ended 2010-02-05 10:30:31 --> Task #1905 finished but not resubmitted. [ERROR] And I send you too the content of error file: 2010-02-05 10:30:31 --> Exception caught: string index out of range
Thanks in advance ________________________________ De: Tibor Simko [mailto:[email protected]] Enviado el: dom 07/02/2010 23:47 Para: LORENTE MARTINEZ, MAGDALENA CC: project-cdsware-users (CDS Invenio users) Asunto: Re: problems with bibindex On Mon, 01 Feb 2010, LORENTE MARTINEZ, MAGDALENA wrote: > My cdsinvenio installation was running ok but suddenly bibindex > process shows an error: idxWORD09F flush ended. `flush ended' is not an error message, but the last good output message. In order to see the error message, when you are in the bibsched monitor, you can press `l' to inspect the task output log, and `L' to inspect the task error log. Alternatively, for a task number 1858, just check out the files /opt/cds-invenio/var/log/bibsched_task_1858.{log,err}. If you don't find anything interesting there, then please also look at the generic invenio log files at /opt/cds-invenio/var/log/invenio.{log,err}. Best regards -- Tibor Simko ** CERN Document Server ** <http://cds.cern.ch/>
