ERROR in log: RecordFormatException: Unable to construct

[isaac737]

Howdy all,

Well we have successfully migrated alf_data from the local Red Hat Enterprise 5 server running Alfresco Community 3.0!

Yee haw…    it was a bit painful but Alfresco is back up and the Content Store has plenty of room now.

However, we are seeing a few of these errors in our alfresco.log file:

:?:

[Wed Jul 07 | root@babbage:~]$ tail -f /usr/share/apache-tomcat-6.0.20/bin/alfresco.log
12:06:27,664 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance
12:06:30,428 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance
12:06:31,523 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance
12:06:31,618 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance
12:06:33,144 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance
12:06:33,978 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance
12:06:35,090 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance
12:06:38,615 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance
12:14:37,784 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance
12:14:39,248 ERROR [org.alfresco.repo.content.transform.PoiHssfContentTransformer] org.apache.poi.hssf.record.RecordFormatException: Unable to construct record instance

They have happened 214 times since the server fired up about 3 hours ago:

[Wed Jul 07 | root@babbage:~]$ grep PoiHssfContentTransformer /usr/share/apache-tomcat-6.0.20/bin/alfresco.log | wc -l
214
[Wed Jul 07 | root@babbage:~]$

I guess our questions are what exactly are these?  Do we need to worry about them and will they somehow corrupt our Alfresco Document Repository? 

We did some research and they appear to be some sort of background process that Alfresco runs periodically to transform docs to other formats?  But why would Alfresco be doing this?  We want a .PDF to stay a .PDF and and .XLS to stay an .XLS

Any help you guys could provide would be MUCH appreciated as this has been somewhat StResSSFuL. 

Thanks a ton.

Cheers,
jg

[mrogers]

Although you may not want to transform your content, alfresco will transform the context to text so it can be added to the full text index.

In this case there seems to be a problem with transforming some of your excel files.    (HSSF stands for "Horrible Spreadsheet Format" in the POI library).

If you can identify which files are causing problems then please raise an issue in Alfresco's JIRA or Apache POI's bug tracking system.

The good news is that except from full text searching of these few files, everything else is O.K.

[isaac737]

Hi Mrogers and everyone else!

Dude, thanks so much for writing back.  That makes us breathe somewhat of a sigh of relief.

Still, we are somewhat concerned about which Excel files aren't being transformed and added to the Full Search Index.   :shock:

Are these files ok- like can people upload and download them?  Just not searchable from the Index search from within Alfresco?

If not searchable, that is bad but not the end of the world.  We are happy to just have the DocRep moved to our new NAS. 

Lastly, how can we identify these files?  Are there some clues to look for in either alfresco.log or catalina.out?

- Jg

p.s. Will installing Open Office on this RHEL 5 Server help with the transforms?