Intelligent file pre-fetch based on access patterns

摘要

Files and/or file updates can be pre-fetched from a server to reduce file access periods as well as shift network usage from high traffic periods to low traffic periods. For example, a file that was previously expelled from a cache can be pre-fetched from the server prior to a future period in which a client is likely to re-access the file. As another example, a file update for a file that remains stored in the cache can be pre-fetched from the server prior to a future period in which a client is likely to re-access the file. The file update may then be used to update the stored file. The file and/or updated-file can then be provided directly from the cache to the client when the client re-accesses the file.

主权项

1. A method for efficient content access, the method comprising:
fetching a file that is editable from a server in response to an individual client accessing the file, wherein the file is stored in a cache memory upon being fetched from the server; expelling the file from the cache memory after the individual client stops accessing the file; assigning an access ranking to the file based on how frequently the file was accessed by the individual client, the file being assigned a higher access ranking when the file is accessed more frequently by the individual client than when the file is accessed less frequently by the individual client; and pre-fetching the file from the server without the individual client attempting to re-access the file when the access ranking assigned to the file is one of the N highest access rankings amongst access rankings assigned to a set of editable files commonly accessed by the individual client, where N is an integer greater than or equal to one, the file being pre-fetched from the server after having been expelled from the cache memory, wherein the file is re-stored in the cache memory upon being pre-fetched from the server.