Issue Summary

Since the introduction of last change to the AnalyticsRepositoryDeletedEvent and AnalyticsRepositoryModifiedEvent some instance are returning Server unreachable error when these events are generated such as deleting or updating a repository.
 
This can be seen in the profiler logs where findByRepository takes a long time to complete

[154156.7ms] - Page com.atlassian.stash.internal.repository.RepositoryActivityDao.findByRepository(int,PageRequest) 

This is reproducible on Data Center: yes

Steps to Reproduce

1. Delete a repository
2. Whilst that repository is deleting attempt to update the name of a different repo

Expected Results

Users are able to update repositories whilst another is being deleted.

Actual Results

Server unreachable error is caused by a timeout when trying to update other repositories while waiting for `findByRepository` to complete.

Workaround

Currently there is no known workaround for this behavior. A workaround will be added here when available