| 摘要 |
A method and apparatus for performing template-based prefix caching advantageously identifies common prefixes (i.e., initial video segments) in video titles, stores common prefixes only once in a prefix cache, and uses these common prefixes when serving requests for video content. This advantageously enables a prefix cache to scale to a large number of video titles, since the cache stores each common prefix only once. A new video title that uses an already existing prefix may be advantageously added without requiring additional storage in the prefix cache. Template-based prefix caching also advantageously reduces the bandwidth required to distribute prefixes when new titles are ingested into the system—if the required template is already available in the prefix cache, prefix-caching is enabled instantly for this title and no additional bandwidth is required to distribute the prefix.
|
