Systems and methods for the efficient reading of data in a server system

A server comprising:

• mass storage configured to store a plurality of files of a streamable format;
• a shared multifile segment (SMS) cache coupled to the mass storage, where the SMS cache is configured to store a plurality of segments of data from the mass storage, where a segment in the SMS cache is accessed by the CPU to provide data to one or more clients;
• and
• a CPU configured to provide a stream of data to a client over a network, where the CPU is configured to retrieve data for the stream of data from the SMS cache, where an identifiable portion of a file format is mapped in virtual memory and can be retrieved from the SMS cache, where another portion of the file format is not mapped in virtual memory but can be retrieved from the SMS cache.

A method of managing a cache comprising:

• storing segments of data in the cache, where the segments are copied from mass storage, where the segments correspond to a first type of data or to a second type of data;
• forming long-term links to segments corresponding to the first type of data, where the long-term links persist even when data is not being extracted from the corresponding segments;
• forming short-life links to segments corresponding to the second type of data, where short-life links are formed when data from the segments from the cache are accessed, where a short-life link is automatically removed when data is no longer extracted from the corresponding segments;
• and
• inhibiting a release of a segment in the cache when a link to the segment is present.

A computer-readable medium having computer-executable instructions for managing a cache comprising:

• a module adapted to store segments of data in the cache, where the segments are copied from mass storage, where the segments correspond to at least two types of data termed a first type of data and a second type of data, wherein the segments of data in the cache can be shared among a plurality of clients;
• a module adapted to form long-term links, where the long-term links persist when data is not extracted from the corresponding segments;
• a module adapted to form short-life links to segments corresponding to the second type of data, where short-life links are formed when data from the segments from the cache are streamed to a client by a CPU, where a short-life link is removed when data is no longer extracted from the corresponding segments;
• and
• a module adapted to inhibit a release of a segment in the cache when a link to the segment is present.

A method of inhibiting release of segments in a cache, the method comprising:

• receiving a request for data, where the requested data corresponds to a type of data that is selected for inhibited release;
• determining whether the data is physically available in a first segment of the cache;
• releasing ownership of a first page fragment object when the data is not available in the cache such that release of the first segment is inhibited, where the first page fragment object is linked to the first segment and is used to retrieve data from the first segment;
• and
• requesting ownership of a second page fragment object such that the requested data is available to be retrieved from a second segment in the cache when the data is not physically available in the first segment of the cache.

A cache management system comprising:

• a means for receiving a request for data, where the requested data corresponds to a type of data that is selected for inhibited release;
• a means for determining whether the data is physically available in a first segment of the cache;
• a means for releasing ownership of a first page fragment object when the data is not available in the cache such that release of the first segment is inhibited, where the first page fragment object is linked to the first segment and is used to retrieve data from the first segment;
• and
• a means for requesting ownership of a second page fragment object such that the requested data is available to be retrieved from a second segment in the cache when the data is not physically available in the first segment of the cache.

A computer-readable medium having computer-executable instructions for managing a cache comprising:

• a module adapted to receive a request for data, where the requested data corresponds to a type of data that is selected for inhibited release;
• a module adapted to determine whether the data is physically available in a first segment of the cache;
• a module adapted to release ownership of a first page fragment object when the data is not available in the cache such that release of the first segment is inhibited, where the first page fragment object is linked to the first segment and is used to retrieve data from the first segment;
• and
• a module adapted to request ownership of a second page fragment object such that the requested data is available to be retrieved from a second segment in the cache when the data is not physically available in the first segment of the cache.

A method of retrieving data for streaming of media presentations to a plurality of clients, the method comprising:

• receiving a request for a media presentation;
• selecting media tracks for the media presentation;
• maintaining virtual memory address spaces for selected child atoms of sample table atoms;
• receiving a request for data for the selected child atoms;
• mapping virtual memory addresses for the selected child atoms to real addresses for a cache when data is requested;
• and
• using the mapped addresses to retrieve the requested data from the cache.

A computer-readable medium having computer-executable instructions for retrieving data for streaming of media presentations to a plurality of clients, the computer-readable medium comprising:

• a module adapted to receive a request for a media presentation;
• a module adapted to select media tracks for the media presentation;
• a module adapted to maintain virtual memory address spaces for selected child atoms of sample table atoms;
• a module adapted to receive a request for data for the selected child atoms;
• a module adapted to map virtual memory addresses for the selected child atoms to real addresses for a cache when data is requested;
• and
• a module adapted to use the mapped addresses to retrieve the requested data from the cache.