Systems and methods for compression of data for block mode access storage
AssigneeInternational Business Machines Corporation
InventorsJonathan AMIT, Noah AMIT, Nadav KEDEM
Life of this patent
Estimating the term…
Abstract
Systems and methods for creating, reading, and writing compressed data for use with a block mode access storage. The compressed data are packed into plurality of compressed units and stored in a storage logical unit (LU). One or more corresponding compressed units may be read and/or updated with no need of restoring the entire storage logical unit while maintaining de-fragmented structure of the LU.
Claims28
For use with a block mode access storage protocol, a method of writing data at a given range to data stored as compressed data in at least one storage logical unit, wherein said compressed data are packed into one or more compressed sections corresponding to fixed-size portions of raw data (clusters) virtually accommodated in a virtual logical unit, the method comprising:
- a) determining a first cluster to be updated thereby constituting the original cluster;
- b) determining a serial number of corresponding original compressed section;
- c) restoring the data corresponding to said original cluster;
- d) calculating an offset of the updating data within said cluster and update at the given data range;
- e) compressing the updated cluster into an updated compressed section;
- and
- f) overwriting said original compressed section with updated compressed section.
A system for use with a block mode access storage protocol, comprising:
- a processor;
- and
- memory coupled to the processor, wherein said memory is configured to store code for executing a method of writing data at a given range to data stored as compressed data in at least one storage logical unit, wherein:
- said compressed data are packed into one or more compressed sections corresponding to fixed-size portions of raw data (clusters) virtually accommodated in a virtual logical unit, and
- said code, when executed by said processor, causes said processor to:
- a) determine a first cluster to be updated thereby constituting the original cluster,
- b) determine a serial number of corresponding original compressed section,
- c) restore the data corresponding to said original cluster,
- d) calculate an offset of the updating data within said cluster and update at the given data range,
- e) compress the updated cluster into an updated compressed section;
- and
- f) overwrite said original compressed section with updated compressed section.
A computer storage medium for use with a block mode access storage protocol comprising a computer program product method for writing data at a given range to data stored as compressed data in at least one storage logical unit, wherein said compressed data are packed into one or more compressed sections corresponding to fixed-size portions of raw data (clusters) virtually accommodated in a virtual logical unit, the computer storage medium comprising:
- a) code for determining first cluster to be updated thereby constituting the original cluster;
- b) code for determining a serial number of corresponding original compressed section;
- c) code for restoring the data corresponding to said original cluster;
- d) code for calculating an offset of the updating data within said cluster and facilitating the update at the given data range;
- e) code for compressing the updated cluster into an updated compressed section;
- and
- f) code for overwriting said original compressed section with updated compressed section.
Dependencies are inferred from the claim text.