We just completed a PoC with JDG and one of the tests was with a larger distributed cache (14GB). I found myself endlessly looking through the .XSD files (JDG_HOME/doc/schema) to find all the timeout values and then just playing trial or error to find an appropriate value for the timeouts. It took a couple of days of trial and error to get values that were appropriate. I felt like the documentation was really lagging here, since there really were not any best practices or guidance around what to modify the timeout values to be from their defaults for larger cache values. I realize that this may vary from network to network, but a starting point of at least what timeout values to change would be very helpful. I had to find each timeout value and change it separately. There really is not a single place in the documentation to go to see what values you need to modify for a larger cache (repl or dist).
+1 to change the default value.
This bug's contents/fixes are now available on https://access.redhat.com/documentation/en-US/Red_Hat_JBoss_Data_Grid/6.3/