At least one node in the cluster cannot reserve the required amount of memory. This might be caused by pinned data in the cache.
Check for events in the event log. Follow the fix procedures to resolve the problem.