Thank you for your reply, Alex and Thomas.
You are right, the cost is not expensive in common especially in case of the
datastore in a intranet.
Actually my cluster servers will be in different countries, and binary file
are more than several hundred Mb size. In each country, there is a JR
clustering instance with its own shared datastore, more than 90% of the
datastore files content will not be visited from the other country, only the
metadata is needed. In this case, I want only the less than 10% files to be
synchronized on demand. Maybe this is out range of jackrabbit clustering
functionality. It might be a architecture issue.
However I am very appreciate for your help. Thanks!
--
View this message in context:
http://www.nabble.com/Could-jackrabbit-metadata-and-binary-file-separately-synchronize-on-demand--tp17552540p17592812.html
Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
