Emc Corporation

Ugur Sezer

Sorin Faibish

Wayne William Duso

Seyfullah H Oguz

Ivan Bassov

Peter C Bixby

John Forecast

Gurjeet S Arora

The version set backup and restore facility responds to a version set backup request by backing up multiple snapshot copies of a storage object concurrently from a storage array to backup storage media. The version set backup and restore facility responds to a version set restore request by restoring a plurality of snapshot copies of the storage object concurrently from the backup storage media to the storage array. The on-tape backup image of the version set includes variable-length extents of data for each of the multiple snapshot copies of the storage object. The variable-length extents of data for each of the snapshot copies of the storage object are grouped together and ordered in the on-tape backup image by increasing or decreasing logical block address.

Efficient backup and restore of storage objects in a version set

Access to movies ranging from very popular movies to unpopular movies is managed by configuring sets of data movers for associated movie rankings, reserving data mover local cache resources for the most popular movies, reserving a certain number of streams for popular movies, negotiating with a client for selection of available movie titles during peak demand when resources are not available to start any freely-selected movie in disk storage, and managing disk bandwidth and primary and local cache memory and bandwidth resources for popular and unpopular movies. The assignment of resources to movie rankings may remain the same while the rankings of the movies are adjusted, for example, during off-peak hours. A movie locked in primary cache and providing a source for servicing a number of video streams may be demoted from primary cache to disk in favor of servicing one or more streams of a higher-ranking movie.

Video file server cache management using movie ratings for reservation of memory and bandwidth resources

Edges are detected in block coded video by a threshold comparison upon the lengths of variable-length codes used for encoding the differential DC coefficients of the pixel blocks. A thinning filter compares the code lengths of the differential DC coefficients of adjacent blocks in order to retain the edge indications of more significant edges and to exclude the edge indications of less significant edges. The edge indications can be split into substantially independent channels for luminance or chrominance, for edges having positive or negative horizontal gradient components, and for edges having positive or negative vertical gradient components. The edge indications for successive frames in an MPEG sequence are compared to each other in various ways in order to detect scene changes.

Edge detection based on variable-length codes of block coded video

The (run, level) pairs in an original series are inspected to determine whether or not modification of the (run, level) pairs would produce a desirable decrease in a number of bits required for variable-length encoding of the series of (run, level) pairs, despite introduction of noise. If so, the (run, level) pairs are modified prior to variable-length encoding. For example, a (run, level) pair of (M, N) is modified by substitution of a first (run, level) pair of (M&#x2212;1, 1) immediately followed by a second (run, level) pair of (0, N). A lookup table or testing of predetermined ranges of run length and level magnitude provides a fast determination of whether or not to modify a (run, level) pair. The decoder can be programmed to reduce the noise introduced by this process by recognizing and rejecting (run, level) pairs that are likely to have been inserted during the encoding process.

Insertion of noise for reduction in the number of bits for variable-length coding of (run, level) pairs

Transform coefficients for blocks of pixels in an original picture are quantized to produce respective sets of quantization indices for the blocks of pixels. The quantization indices for at least some of the blocks are produced by using a quantization step size that is not uniform within each block. Largest magnitude quantization indices are selected from the respective sets of quantization indices for (run, level) encoding to produce the (run, level) encoded picture. For example, MPEG-2 coded video includes a set of non-zero AC discrete cosine transform (DCT) coefficients for 8&#xd7;8 blocks of pixels. For scaling the MPEG-2 coded video, non-zero AC DCT coefficients are removed from the MPEG-2 coded video to produce reduced-quality MPEG-2 coded video that includes no more than a selected number of largest magnitude quantization indices for the non-zero AC DCT coefficients for each 8&#xd7;8 block.

Growing community of inventors

Shrewsbury, MA, United States of America

Ugur Sezer