When SVN operations timeout after x tries application administrators should be notified

In reference to FE-3253 when SVN operations timeout Fisheye appears to go into a loop of:

1. Execute SVN command
2. Wait to complete
3. Timeout
4. Pause for a bit
5. Goto 1

When faced with a SVN operation that takes permanently longer than <insert timeout value here> after a few tries it would be good if the repository was put into (potentially) a stopped state so that it stops attempting to conduct an operation which can not possibly complete (& it should know this since it's already tried x times)

At this point the application should inform a fisheye-administrator who should then investigate how long the operation may actually take & then take the appropriate action to extend the timeout value or reconsider the position of the repository if it's located too far away to be effective.

At the moment without specific vigilance & intervention we've noticed that Fisheye will continue to try forever & with each iteration it appears to consume more heap space until eventually Fisheye runs out of memory.

Given that there are limited indexing threads (though definable) we consider it important to know if the repo you've just configured will actually never be able to index to a natural conclusion