Cache Hit Rate is an important
indicator
to measure the efficiency and performance of read cache. It has nothing to do with write caching, nor does it reflect write performance. The cache hit rate is calculated according to the following formula:
Cache Hit Rate = Cached Read / Total Read * 100%.
Obviously, the more data is read from the cache, the higher the cache hit rate.
When the data is requested for the first time, PrimoCache will read the data from the disk and store it in the cache, because the data is not in the cache at this time. At this point, the hit rate drops because "Total Read" increases but "Cached Read" does not. So if Windows or programs frequently request new data that has not been read, the hit rate will not be too high.
When requesting to read the data again, PrimoCache will directly read from the cache, and the hit rate will increase at this time. Of course, this is the most ideal situation. In fact, since the cache space is always limited, when the cache is full of data, PrimoCache will discard some old cache data according to the cache replacement algorithm in order to cache new data. At this time, if Windows requests to read old data that has been discarded, PrimoCache needs to read from the disk again. If this happens, the cache hit rate will drop.
It can be seen from the above that to improve the cache hit rate, the cache size can be increased, because the larger the cache, the more data can be stored and the less data will be discarded.
In addition, reducing the number of logical volumes under the cache task can also help improve the cache hit rate. The cache space under a cache task is shared with all logical volumes. Therefore, the more volumes are accelerated, then when the cache space is full, reading data on any volume will cause old data to be discarded, affecting the hit rate of other volumes. So when creating a cache task, you can
accelerate
only the most needed logical volumes instead of all volumes. If
resources are sufficient, independent cache tasks can be created for
different logical volumes, so that the respective cache spaces will not
affect each other.
In addition, enabling the Prefetch function in PrimoCache also helps to improve the cache hit rate. Since the data in the RAM cache is not persistent after shutdown, Prefetch can pre-read the data into the RAM cache so that when Windows first requests to read the data, it can be read from the cache.