We have received a passivation and and activation deadlock error during the time PR_PAGE_STORE was being cleaned up by SystemCleaner. Is there any problems or cautions with switching PR_PAGE_STORE and PR_SYS_CONTEXT to use row level locking?
We were having a problem with data purging. However, we staggered the SystemCleaner agent on our nodes (suggested in an SR) and after a few days, the errors went away.
We did, however, have a couple of times when end users deadlocked on passivation/activation. While SystemCleaner is not taking very long to run now, I would also like to switch to row level locking to help ensure this does not occur in the future.