Search doesn't work (5.1.8)

[Bart]

Can we merge this thread with this one: http://forum.openkm.com/viewtopic.php?f=3&t=5578 ??

[Bart]

Continuing here. see also http://forum.openkm.com/viewtopic.php?f=4&t=5572

I just installed OpenKm 5.1.7 (previous release) and tested with the same file. Search now works! So we definitly know it's a 5.18 issue..

What's is the best way to start looking for the problem? I have increased the loglevel and added the category com.openkm.servlet.frontend. But the amount of logging is HUGE!

[yfzhou72]

I've got the same issue. I'm running a fresh intallation of 5.1.8 on Windows 2003, JRE 7u2. Everything get ok except no results when search content. How OpenKM search engineer working, and is there any log to track what happening?

BR,

Sam

[pavila]

Posts merged as requested.

[pavila]

I think that text extractors are disabled in repository.xml, please attach this file. Anyway to enable text extraction, go to Administration > Configuration and check the experimental.text.extraction property.

[yfzhou72]

I've just enable the "experimental.text.extraction", and restart the application. But seems not work still. Attached please find my repository.xml.

[yfzhou72]

Just confirmed, after enable "experimental.text.extraction", and restart the application, and re-upload the files, the search content function works! Thanks a lot to pavila.
I've test word file, text file and PDf, all works fine.

BR,

Sam

[pavila]

Nice to read

[icek]

Hi Pavila,

I have just installed the 6.14 OpenKM Pro Trial version for my company. I would like to make a demo to my bosses.
The issue is that the search function doesn't work for the content, within documents (txt, doc...). It works fine for the name.
There is no experimental.text.extraction option in Administration-> Config.

[jllort]

Take in mind documents go into text extraction queue and processed in a 5 minutes period. That mean documents content are not indexed just in time ( can be done but is not the default procedure ) go into queue and are sequentially processed. I suggest go to Administration -> Stats -> Indexing queue to see if still you got pending documents to be indexed there. When documents are indexed extracted content value can be shown in database go to Administration -> Utilities -> Database query and execute select * from OKM_NODE_DOCUMENT ( you can see a column what indicate if text has already been extracted = T and the content ). With this information we'll got more information about what happens with your documents. You can force indexing from crontab task and also indexing more document per cycle with Administration -> Configuration parameters -> managed -> property managed.text.extraction.batch

[pavila]

If you want to force the text extraction, go to Administration > Crontab and execute the Text Extractor Worker task.

[jan.sladek]

Hi,

I have just installed the openKM 6.3.0. on Ubuntu 14.04 and have a problem with search engine. The Default OkmAdmin user can search for everything wihtout problem, while custom created users with custom created profile can't search for the documents they uploaded. Either searching by name or contnent, the text, PDF, Openoffice or MS office documents can't be found. I even tried to assign the Default user profile as Admin to the custome user, but there are no results for the search query. What could be the settings/permissions that restrict users from getting the proper results of search query?

thanks

Jan

[jllort]

Please jan, add new post about it, because has not relation with search in 5.1.8 or 5.1.X version. Merge questions on same post will generate confusion to other readers.