News Column

Patent Issued for System and Method for Information Handling System Redundant Storage Rebuild

May 27, 2014

By a News Reporter-Staff News Editor at Information Technology Newsweekly -- A patent by the inventors Nelogal, Chandrashekar (Round Rock, TX); Le, Bang Kim (Broomfield, CO); Yu, Frank Widjaja (Austin, TX), filed on September 27, 2010, was published online on May 13, 2014, according to news reporting originating from Alexandria, Virginia, by VerticalNews correspondents.

Patent number 8726070 is assigned to Dell Products L.P. (Round Rock, TX).

The following quote was obtained by the news editors from the background information supplied by the inventors: "The present invention relates in general to the field of information handling system storage, and more particularly to a system and method for information handling system redundant storage rebuild.

"As the value and use of information continues to increase, individuals and businesses seek additional ways to process and store information. One option available to users is information handling systems. An information handling system generally processes, compiles, stores, and/or communicates information or data for business, personal, or other purposes thereby allowing users to take advantage of the value of the information. Because technology and information handling needs and requirements vary between different users or applications, information handling systems may also vary regarding what information is handled, how the information is handled, how much information is processed, stored, or communicated, and how quickly and efficiently the information may be processed, stored, or communicated. The variations in information handling systems allow for information handling systems to be general or configured for a specific user or specific use such as financial transaction processing, airline reservations, enterprise data storage, or global communications. In addition, information handling systems may include a variety of hardware and software components that may be configured to process, store, and communicate information and may include one or more computer systems, data storage systems, and networking systems.

"As information handling systems have become increasingly prevalent in enterprises and homes, they have generated tremendous amounts of information for storage and subsequent access. For example, businesses maintain customer, accounting, inventory, research, product design and other critical types of information. As another example, individuals maintain personal financial information, pictures, videos and other important personal information. In order to protect important information from disappearing in the event of a storage failure, enterprises and individuals often use redundant storage techniques. One such category of redundant storage techniques is the use of a redundant array of independent disks (RAID). Although a variety of RAID configurations exists, a typical RAID array combines multiple hard disk drives into LUNs that distributes a copy of information stored on any one hard disk drive to at least one other hard disk drive of the array so that a redundant copy of all information in the array exists for use in the event of a hard disk drive failure. To accomplish such redundancy, each hard disk drive is divided into fixed-size stripe units made of several disk sectors, such as units of 8 KB to 256 KB of storage. A parity map maintains a relationship between these units to track where the units are stored on different hard disk drives. After a failure occurs at a hard disk drive, the information stored on the failed hard disk drive is regenerated from the parity data and written to a replacement hard disk drive.

"One difficulty with recreating a failed hard disk drive on replacement hard disk drive is that the data reconstruction operation, also known as an XOR operation, is time consuming and must include several I/O's made to all disk units of the array during the rebuild process. Since the rebuild process for servicing Input/Output (I/O) uses both a rebuild pointer and rebuild status for each region, these parameters must be maintained during the rebuild. During a rebuild process, a RAID controller performs the data reconstruction operation by rebuilding each logical unit starting with a first logical block of the failed hard disk drive and proceeds sequentially to the last logical block with I/O regeneration happening at the granularity of stripe element size. The result of a conventional rebuild process is an identical drive, however reconstructing the data with XOR operations is very time consuming and the overhead associated with the sequential XOR operations can slow overall system performance. For example, one typical 1TB replacement drive re-creation took 16 hours to complete. Until the failed drive is re-created, failure of another hard disk drive within the same array can result in permanent loss of information. Some steps can result in reduced rebuild time, such as avoiding rebuild of portions of the failed drive that did not save data, however, performing the XOR calculations to rebuild the remaining portions of the drive still consumes time and resources.

"Due to increased time to rebuild, the LUN is susceptible to subsequent hard disk drive failure. Hence there is a need to reduce the overall time taken to rebuild a failed drive of a RAID array at a replacement drive."

In addition to the background information obtained for this patent, VerticalNews journalists also obtained the inventors' summary information for this patent: "Therefore a need has arisen for a system and method which rebuilds redundant storage using less time and resources.

"In accordance with the present invention, a system and method are provided which substantially reduce the disadvantages and problems associated with previous methods and systems for rebuilding redundant storage. Instead of the traditional method where data is reconstructed sequentially from the beginning to the end of the hard drive using parity information, whenever possible old data is retrieved directly from a partially failed disk to avoid costly reconstruction and reconstruction is performed out of order based upon concurrent host I/O directed to the failed storage device during reconstruction. Regions associated with an I/O are given an increased priority for reconstruction so that I/O can be executed directly on a replacement disk to avoid the time consuming parity calculations based on data of other members of the array. A failed storage device is reconstructed out of order based upon recoverable information retrieved from the failed storage device and I/O directed to the failed storage device during reconstruction. Regions associated with an I/O are given an increased priority for reconstruction so that I/O are executed on reconstructed regions of a replacement storage device.

"More specifically, an information handling system has plural processing components that cooperate to generate information for storage, such as a CPU, RAM and a chipset. The processing components send I/O commands to a RAID controller for execution at plural storage devices of a RAID array, such as write and read commands. The RAID controller stores information on the plural storage devices with redundant information, such as parity information to one or more strips within each stripe of the array. If a storage device fails, a reconstruction module of the RAID controller retrieves redundant information from functioning storage devices to reconstruct the failed storage device. A priority module adjusts the order in which reconstruction occurs to support a more efficient reconstruction. The priority module analyzes the failed storage device to recover information available from the failed storage device for copying to the replacement storage device. The analysis returns whether or not the disk drive is partially failed and some data can still be retrieved from it. Based on this information, the reconstruction operation makes decisions to retrieve old data either by quickly reading from partial failed disk or following the well known data reconstruction method based on the RAID level of the array. The priority module tracks regions of the replacement storage device that are reconstructed so that I/O directed to reconstructed regions are handled by the replacement storage device while other regions are reconstructed. If the priority module detects an I/O associated with a region of the failed storage device that is not yet reconstructed on the replacement storage device, the priority module increases the priority of reconstruction of that region and queues the I/O until the reconstruction is performed so that the I/O is executed at the replacement storage device. Once the region is completely reconstructed, a status is updated and future I/O to that region are executed at the replacement disk.

"The present invention provides a number of important technical advantages. One example of an important technical advantage is that a redundant storage device is rebuilt in a rapid manner with reduced resources. In one example embodiment, checking a failed storage device to determine whether recoverable information exists helps to reduce rebuild time and resources by avoiding the need to perform parity data calculations, such as XOR operations, at all. Where parity operations are required, the time needed to perform the parity operations is reduced by performing the operations out of order, which avoids rebuilding the same region more than once as might occur with the traditional method. Prioritizing the order for rebuilding based upon current I/O operations means that an I/O, such as a write, received from a host is always executed in a non-reconstructed mode and thus latency at most would only incur on the first I/O accessing the non-reconstructed region. In one example embodiment of a partial failed drive, a rebuild time for a RAID hard disk drive storage device was reduced from 16 hours down to 2 hours."

URL and more information on this patent, see: Nelogal, Chandrashekar; Le, Bang Kim; Yu, Frank Widjaja. System and Method for Information Handling System Redundant Storage Rebuild. U.S. Patent Number 8726070, filed September 27, 2010, and published online on May 13, 2014. Patent URL:

Keywords for this news article include: Dell Products L.P., Information Technology, Information and Data Storage, Information and Data Loss and Recovery.

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: Information Technology Newsweekly

Story Tools Facebook Linkedin Twitter RSS Feed Email Alerts & Newsletters