News Column

Patent Issued for Methods and Apparatus for Improved Raid Parity Computation in a Storage Controller

July 3, 2014



By a News Reporter-Staff News Editor at Computer Weekly News -- A patent by the inventors Hall, Randy K. (Boulder, CO); Gates, Dennis E. (Wichita, KS); Sterns, Randolph W (Arvada, CO); Kloeppner, John R. (Buhler, KS); El-Batal, Mohamad H. (Boulder, CO), filed on October 12, 2011, was published online on June 17, 2014, according to news reporting originating from Alexandria, Virginia, by VerticalNews correspondents.

Patent number 8756371 is assigned to LSI Corporation (Milpitas, CA).

The following quote was obtained by the news editors from the background information supplied by the inventors: "The invention relates generally to storage systems and more specifically relates to methods and structure for improved performance in computing parity in a RAID storage controller by using both hardware parity computation assist circuitry and programmed instructions for parity computation.

"In RAID (Redundant Array of Independent Disks) storage systems, logical volumes are defined and managed. A logical volume comprises portions of two or more disk drives and stores redundancy information along with user data provided by attached host systems. The redundancy information is generated and stored in conjunction with stored data such that failure of a single disk drive (or two disk drives in some embodiments) does not preclude ongoing access to the data on the logical volume. Rather, the logical volume may continue operation until such time as the failed drive is replaced and the full redundancy features are restored (a 'hot swap' operation and a 'rebuild' operation).

"Various 'levels' of RAID storage management are standardized in the storage industry. In RAID level 5, user data and redundancy information is distributed ('striped') over a plurality of storage devices (at least 3 devices). In RAID level 5, the redundancy information is the exclusive-OR ('XOR') sum of the user data. Thus, a 'stripe' of data in RAID level 5 management includes a plurality of blocks of user data and a corresponding parity block that is the XOR sum of the user data in the related blocks of the stripe. In like manner, RAID level 6 defines a stripe of data as a plurality of blocks of user data and two blocks of redundancy information--typically a first parity block that is the XOR sum or the other blocks and a second block that may be also XOR parity, Galois field accumulation (also referred to as Galois field multiplication or 'GFM'), or other redundancy computations. RAID level 6 can therefore keep a logical volume operating even in the event of two disk drive failures of the logical volume.

"Computation of the redundancy block (or blocks) can impose a notable computational load on a general purpose processor of the storage controller. For this reason, a number of present storage controllers replace the software ('firmware') computation of a redundancy block (or blocks) with a special purpose circuit designed to assist the general purpose processor by off-loading the redundancy computations to a specially designed circuit for such computation. However, even such special purpose processors can become overloaded in a high performance storage system and/or may be less efficient for certain types of data or certain patterns of user data access.

"Thus it is an ongoing challenge to efficiently compute redundancy information used in RAID storage systems and in particular in RAID levels 5 and 6 storage management."

In addition to the background information obtained for this patent, VerticalNews journalists also obtained the inventors' summary information for this patent: "The present invention solves the above and other problems, thereby advancing the state of the useful arts, by providing methods and apparatus for combining a firmware/software element ('firmware parity engine' or 'FPE') for generating redundancy information with a custom logic circuit ('hardware parity engine' or 'HPE') designed to generate redundancy information. A scheduler element operable on a processor of a storage controller along with the FPE determines which of the FPE and HPE is best suited to rapidly complete a new redundancy computation operation. The scheduler then activates an idle component (FPE or HPE) to process the new operation or queues the new operation for eventual processing of the new operation.

"In one aspect hereof, a storage controller is provided comprising a programmable processor programmed to process I/O requests from attached systems and a firmware redundancy computation element operable on the programmable processor of the storage controller. The firmware redundancy computation element is operable to generate redundancy information from data applied to the firmware redundancy computation element. The controller further comprises a redundancy computation circuit coupled with the programmable processor and operable to generate redundancy information from data applied to the redundancy computation circuit. The controller also comprises a scheduler element operable on the programmable processor. The scheduler element is operable to selectively utilize either the firmware redundancy computation element or the redundancy computation circuit for generating redundancy information in the storage controller based on present operating conditions of the storage controller. The redundancy information is used by the controller in processing RAID storage management relating to the I/O requests.

"Another aspect hereof provides a method and an associated computer readable medium embodying the method. The method is operable in a Redundant Array of Independent Disks (RAID) storage controller for generating redundancy information. The storage controller has a firmware redundancy computation element operable on a programmable processor of the controller. The storage controller also has a redundancy computation circuit coupled with the programmable processor. The method comprises generating a request to perform a redundancy computation operation based on identified data. The request comprises information identifying the computation to be performed. The method also comprises selecting a selected redundancy computation component from either the firmware redundancy computation element or the redundancy computation circuit to perform the redundancy computation operation based on present operating status of the storage controller and performing the redundancy computation operation using the selected redundancy computation component to generate redundancy information. The method then uses the generated redundancy information for RAID storage management functions of the storage controller."

URL and more information on this patent, see: Hall, Randy K.; Gates, Dennis E.; Sterns, Randolph W; Kloeppner, John R.; El-Batal, Mohamad H.. Methods and Apparatus for Improved Raid Parity Computation in a Storage Controller. U.S. Patent Number 8756371, filed October 12, 2011, and published online on June 17, 2014. Patent URL: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=8756371.PN.&OS=PN/8756371RS=PN/8756371

Keywords for this news article include: Software, LSI Corporation.

Our reports deliver fact-based news of research and discoveries from around the world. Copyright 2014, NewsRx LLC


For more stories covering the world of technology, please see HispanicBusiness' Tech Channel



Source: Computer Weekly News


Story Tools






HispanicBusiness.com Facebook Linkedin Twitter RSS Feed Email Alerts & Newsletters