Three additional asynchronous prefetch statistics are reported in this section:
“APFs Used” reports the number of pages that were brought into the cache by asynchronous prefetch and used during the sample interval. The pages counted for this report may have been brought into cache during the sample interval or by asynchronous prefetch requests that were issued before the sample interval started.
“APF Waits for I/O” reports the number of times a process had to wait for an asynchronous prefetch to complete. This indicates that the prefetch was not issued early enough for the pages to be in cache before the query needed them. It is reasonable to expect some percentage of “APF Waits.” Some reasons that tasks may have to wait are:
The first asynchronous prefetch request for a query is generally included in “APF Waits.”
Each time a sequential scan moves to a new allocation unit and issues prefetch requests, the query must wait until the first I/O completes.
Each time a nonclustered index scan finds a set of qualified rows and issues prefetch requests for the pages, it must wait for the first pages to be returned.
Other factors that can affect “APF Waits for I/O” are the amount of processing that needs to be done on each page, and the speed of the I/O subsystem.
“APF Discards” indicates the number of pages that were read in by asynchronous prefetch and discarded before they were used. A high value for “APFs Discards” may indicate that increasing the size of the buffer pools could help performance, or it may indicate that APF is bringing pages into cache that are not needed by the query.
