This document shows the terms and definitions that are specific to the statistics and performance of a cache task. Statistics are calculated since Windows boot, if you didn't reset the statistics. Otherwise they are calculated since your last reset of the statistics.
Cached Read: Total size in bytes of read operations done on the L1 and the L2 cache. This is a subset of Total Read statistic. The number in parenthesis represents the cache hit rate.
L2Storage Read: Total size in bytes of reads from the level-2 storage. The number in parenthesis represents the level-2 cache hit rate which is calculated from the formula: L2Storage Read / Total Read * 100%.
L2Storage Write: Total size in bytes of writes populating read-data from the underlying disk to the level-2 storage.
Total Write (L1/L2): Total bytes written to level-1/level-2 cache space with respect to write requests.
Total Write (Disk): Total bytes written to the underlying disk. The number in parenthesis is calculated from the formula: Total Write (Done) / Total Write (Req) * 100%.
Urgent Write: Total size in bytes of data written to the disk from the cache by reason of insufficient free cache space. When cache space is almost full, PrimoCache has to urgently write certain amount of deferred data stored in the cache to the disk, thus provides enough cache space for new incoming data.
Normal Write: Total size in bytes of data written to the disk from the cache by reason of timer flush or manual flush operations.
Deferred Blocks: Number of cache blocks which currently stores write-data to be written to the disk. The number in parenthesis is the percentage of deferred blocks out of total cache blocks.
Prefetch: Indicates the prefetch state. The numbers in parenthesis refers to the total bytes prefetched (the number before backslash) and the total bytes requested to be prefetched (the number after backslash).
Free Cache (L1/L2): Size in bytes of currently free L1/L2 cache space for new data. When Free Cache drops to a certain value (32MB), PrimoCache will discard or move around some cache blocks according to the cache algorithm.