Nearstore compression of data in a storage system

A storage server, comprising:

• a processor;
• a network adapter coupled with the processor, wherein the network adapter is configured to receive, from a client system, a request to store a data block, and a request for the data block;
• a compression module coupled with the network adapter, wherein the compression module is configured to:
• compress the data block in a compression group in response to receiving the request, wherein the compression group includes a plurality of logically related compressed data blocks, wherein each of the plurality of logically related compressed data blocks are referenced by an indirect block pointer in a buffer tree,
• search for the compression group delimiter within metadata associated with a predetermined number of data blocks adjacent to the compressed data block to determine whether the data block is within the compression group, wherein the searching is performed before decompressing the compressed data block,
• in response to the request for the data block, decompress a compressed version of the data block in a buffer cache coupled to the compression module;
• and
• a nonvolatile memory coupled with the compression module, wherein the nonvolatile memory is configured to:
• store the data block, wherein a location for the data block is stored in an indirect block in the buffer tree having a root node containing storage server file information, the data block storing actual data that comprises a file;
• and
• store the compression group prior to flushing the compression group to a physical storage device in response to reaching a consistency point.

A storage server, comprising:

• a processor;
• and
• a memory coupled with the processor, wherein the memory contains instructions that, when executed by the processor, cause the storage server to:
• receive a request to store a data block, wherein a location for the data block is stored in an indirect block in a buffer tree having a root node containing storage server file information, the data block storing actual data that comprises a file;
• in response to receiving the request, store a compressed version of the data block in a compression group, wherein the compression group includes a plurality of logically related compressed data blocks, wherein each of the plurality of logically related compressed data blocks are referenced by an indirect block pointer in the buffer tree;
• store the compressed version of the data block in a buffer cache;
• receive, a request for the data block;
• in response to the request for the data block, search for the compression group delimiter within metadata associated with a predetermined number of data blocks adjacent to the compressed data block to determine whether the data block is within the compression group, wherein the searching is performed before decompressing the compressed version of the data block from the buffer cache;
• send the requested data block to a client;
• store the compression group in a nonvolatile memory prior to flushing the compression group to a physical storage device;
• and
• in response to reaching a consistency point, flush the compression group form the nonvolatile memory to the physical storage device.

A computer-implemented method, comprising:

• storing a compressed version of a data block in a compression group in a physical storage device of a storage server, wherein a location for the data block is stored in an indirect block in a buffer tree having a root node containing storage server file information, the data block storing actual data that comprises a file, the compression group including a plurality of logically related compressed data blocks, and wherein each of the plurality of logically related compressed data blocks are referenced by an indirect block pointer in the buffer tree;
• retrieving the compression group from the physical storage device in response to a request for the data block;
• storing the retrieved compression group in a buffer cache of a system memory of the storage server;
• and
• searching for the compression group delimiter within metadata associated with a predetermined number of data blocks adjacent to the compressed data block to determine whether the data block is within the compression group, wherein the searching is performed before decompressing the compressed version of the data block from the buffer cache.