Consider reducing contention for pages by changing index definitions in a way that alters the physical distribution of data across the data and index pages within your table, or modifying your application to reduce contention.
If the average value for WaitTime is high, event 41 may occur because of an Adaptive Server resource shortage, resulting from:
A hash table that is too small for a lock, resulting in very long hash chains that Adaptive Server must search.
An operating system issue during which calls that should be quick are a bottle neck (for example, starting asynchronous I/O, which should return immediately, blocks because of operating system resource limitations.
Extremely high inserts and expanding updates. Page allocations take place frequently, and contention for the allocation page latch results in a high number of Waits. Use dbcc tune(des_greedyalloc) to reduce this contention. For information about latch contention, see Performance and Tuning Series: Monitoring Adaptive Server with sp_sysmon.
