| 摘要 |
A caching manager calculates cache priories for accessed items, including items not currently stored in the cache. The cache manager determines an item's cache priority as a function of the item's access frequency, retrieval cost and size. The cache manager dynamically updates cache priorities as items are accessed. When the cache is full and a non-cached item is accessed, the cache manager uses cache priorities to determine whether to overwrite a cached item with the non-cached accessed item. If the accessed item has a lower cache priority than all of the items in the cache, then the non-cached item is not stored in the cache. If any item in the cache has a lower cache priority than the accessed item, then the cache manager stores the accessed item in the cache, overwriting the cached item with the lowest cache priority.
|
