I just installed RhodeCode on our somewhat large HG repo (~12K changesets). The goal is to replace HGWeb which is slow on our repo and does not seem to permit certain operations concurrently, so it locks some users. Am I right that RhodeCode will be faster and permits concurrent access?
With statistics enabled, RhodeCode is really slow and also consumes a lot of ram. In the statistics, it says that ~15% of the statistics are available and it is consuming ~4GB of ram. There is a single user on the repo now since it is only for testing.
Disabling the statistics makes RhodeCode a lot faster. BTW, I did not deploy RhodeCode with Celery, yet.