PROTOCOL FOR IMPROVING REBUILD TIMES OF REDUNDANT ARRAY OF INDEPENDENT DISKS

摘要

A method and system for rebuilding a Redundant Array of Independent Disks ("RAID"), the system comprising a RAID engine comprising one or more processing devices that facilitate packet communications with a plurality of storage devices, the packet communications including Read/Write-Rebuild opcodes and settings that prioritize read and write requests associated with a RAID rebuild of the plurality of storage devices, the Read/Write-Rebuild opcodes including rebuild input/output ("10") read requests that read good portions of data and parity information from a set of the plurality of storage devices, and rebuild 10 write requests that write data to a new one of the plurality storage devices wherein the written data includes data that is reconstructed by the RAID engine based on the good portions of data and the parity information.