jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Lorenzo Dini <Lorenzo.D...@cern.ch>
Subject Re: Some performance questions about Jackrabbit
Date Fri, 01 Feb 2008 18:11:25 GMT

> 3) Export all data from the old Jackrabbit excluding BLOBs of course -
> instead export just the UUID of the BLOB (should be possible - is this
> what you did so far?)

Yes it is possible because the UUID of the BLOB is the UUID of the node 
containing the binary property.

The full URL is something like this

blobs/node UUID containing the bin property divided in 3 directory 

 > 4) Import all data, but instead of importing the BLOB or the UUID of
 > the blob, use the right SHA-1 hash.

How do you do that? The importXML needs to read the lookup table that we 
generated and set the property with DataIdentifier.

It is needed to extend the importXML functionality. In my opinion, the 
best way would be to provide the DataIdentifier in the XML itself 
(instead of the content) if the user chooses not to export binaries. 
Right now the content of the BIN property XML node is not used in case 

I think the usecase of storing all the binaries in a single XML is not 
so common if you have a large storage.

Providing this extra functionality would solve the problem because the 
lookup table can be the XML file itself :-)

In this way the Store conversion tool will just replace the content of 
the XML changing the old DataIdentifier (Node UUID) with the new one 

> Would that make sense?



> Regards,
> Thomas

*Lorenzo Dini*

CERN - European Organization for Nuclear Research
Information Technology Department
CH-1211 Geneva 23

Building 28 - Office 1-007
Phone: +41 (0) 22 7674384
Fax: +41 (0) 22 7668847
E-mail: Lorenzo.Dini@cern.ch

View raw message