Patent number 8819380 is assigned to
The following quote was obtained by the news editors from the background information supplied by the inventors: "At present, hard disk drives are one of the most common forms of electronic data storage devices. Hard disk drives typically contain multiple magnetic disks ('platters') which serve as the storage media. The platters are rotated and magnetic heads write data as a series of magnetic polarity transitions in a plurality of concentric circular tracks on the surfaces of the platters. For example, a track region of uniform magnetic polarity can represent a '0' and a track region containing a transition from one magnetic polarity to the opposite magnetic polarity can represent a '1', where each '0' or '1' constitutes a bit. The magnetic heads can also read stored data by moving into position over particular tracks and detecting these magnetic polarity transitions (or the absence thereof) as the platters rotate beneath.
"Each platter is divided into geometrical sectors. The intersections of geometrical sectors and the concentric circular tracks define data sectors. Each data sector typically consists of 512 bytes, and each byte typically consists of 8 bits. A block, which typically represents the smallest storage unit that can be allocated for storing data, consists of a contiguous group of data sectors. For example, a 4 kilobyte block typically consists of 8 contiguous data sectors (i.e., 512 byte sectors).
"An ever-present concern in hard disk drive technology is adjacent track interference (ATI) and wide area adjacent track erasure (ATE). ATI occurs when a head successively writes data on a particular track of a platter and the magnetic field of the head weakens the magnetic polarity transitions on adjacent tracks, potentially to the extent that the data on the adjacent tracks becomes unreadable. ATE is a similar effect that has the potential to corrupt data on a larger number of adjacent tracks. For example, if data stored on track N is frequently modified (e.g., the head is frequently performing writes to that track region), data located on adjacent tracks N+1 and N-1 that is near the high-write region of track N can be subject to ATI and ATE. As the disparity in write frequencies of data on adjacent tracks becomes greater, the potential for ATI and ATE typically increases. For example, infrequently written data that is adjacent to a track region with a high write frequency is especially susceptible to ATI and ATE.
"A file system enables applications to store and access data on a hard disk drive. To store a file on a hard disk drive using a typical file system, the file system must initially allocate a whole number of free blocks for the file to use. Based on the anticipated size of a file to be stored, an application can issue allocation requests to the file system for the appropriate number of free blocks. Later, if the size of the file increases, the file system can allocate additional free blocks; if the file size is reduced or the file is deleted, the file system can free the file's allocated blocks for allocation to other files. It should be noted that, while each individual block consists of contiguous sectors, a group of free blocks allocated for a particular file need not be contiguous. For example, if a contiguous group of free blocks equal to the requested amount is not available, a single file can be allocated to a group of free blocks that are non-contiguously distributed (i.e., fragmented) over the tracks of a platter or multiple platters within the stack of a hard disk drive.
"A file system's allocation method determines which particular free blocks to allocate for a file. A common allocation method involves giving preference to allocating a contiguous group of free blocks on one of the outermost tracks of a platter. This method can ultimately enable faster access to the file. For example, if all blocks allocated for the file are contiguous on a single track, the entire file can be accessed on a single platter rotation (i.e., reduced rotational latency) and without having to move the heads to multiple tracks (i.e., reduced seek time). Further, this method takes advantage of the higher media transfer rates of outer tracks (i.e., with zone bit recording and constant rotational velocity of platters, the greater linear velocity and quantity of blocks on outer tracks enables more of those blocks to be accessed by a head per unit of time relative to blocks on inner tracks). However, as a consequence of giving preference to free blocks located on outer tracks, these regions will often experience a high number of writes, while inner tracks adjacent to these high-write regions will often experience a lesser number of writes. Accordingly, this method can result in a distribution of allocated blocks that is ripe for ATI and ATE effects.
"One technique to minimize the potential for ATI and ATE involves reducing the number of tracks per inch (TPI) on the platters. By decreasing the TPI, tracks are spaced farther apart, which can help reduce the extent to which the heads' magnetic fields affect the data stored on adjacent tracks. However, decreasing the TPI typically reduces the storage capacity of a platter. Further, decreasing TPI can result in poorer performance due to increased seek times.
"Other techniques to minimize the potential for ATI and ATE involve periodically performing refresh cycles. In a typical refresh cycle, the data on a track that is potentially subject to ATI and ATE is read and then rewritten in the same location. Rewriting the data helps ensure its continued integrity by strengthening the magnetic polarity transitions that may have been weakened by ATI and ATE. The tracks on which to perform refresh cycles can be determined by monitoring the number of write operations on each track and, upon reaching a threshold number of write operations for a given track, performing a refresh cycle on the tracks adjacent to that track. However, performing refresh cycles can negatively impact performance on account of increased seek times and because certain read and write functions cannot be performed during refresh cycles. Further, performing refresh cycles can involve additional head movement, additional platter rotations, and other hard disk drive processes that may increase power consumption."
In addition to the background information obtained for this patent, VerticalNews journalists also obtained the inventors' summary information for this patent: "Embodiments of the present invention provide a system, method, and program product for allocating a block of physical storage space on a write surface of a hard disk drive. A computer system maintains a write count for each block on the hard disk drive, wherein a write count is a cumulative number of write operations performed on a block of storage space. The computer system receives an allocation request for a block of storage space on the hard disk drive. The computer system identifies one or more candidate blocks of storage space on the hard disk drive that can be selected to fulfill the allocation request. The computer system determines an estimated write count, wherein the estimated write count is an estimated number of write operations that will be performed on any candidate block that is subsequently selected from the one or more candidate blocks to fulfill the allocation request. The computer system identifies one or more allocated blocks of storage space on the hard disk drive whose write counts are within a specified number of write operations of the estimated write count. The computer system selects a candidate block from the one or more candidate blocks based, at least in part, on physical proximity of the candidate block to one or more of the one or more allocated blocks of storage space on the hard disk drive whose write counts are within a specified number of write operations of the estimated write count. The computer system allocates the selected candidate block to fulfill the allocation request.
"In certain embodiments of the present invention, the computer system selects a candidate block from the one or more candidate blocks that is located on a track that is adjacent to one or more tracks on which the one or more allocated blocks of storage space whose write counts are within a specified number of write operations of the estimated write count are located."
URL and more information on this patent, see: Patil, Sandeep R.; Ramanathan, Sriram; Shiraguppi, Riyazahamad M.; Trevathan, Matthew B.. Consideration of Adjacent Track Interference and Wide Area Adjacent Track Erasure during Block Allocation. U.S. Patent Number 8819380, filed
Keywords for this news article include:
Our reports deliver fact-based news of research and discoveries from around the world. Copyright 2014, NewsRx LLC
Most Popular Stories
- Stocks Close Out Best Week in Nearly 2 Years
- Microsoft Earnings Drive Stocks Higher
- Do Voters Want Compromise?
- Timeline for New York Ebola Doctor
- 'Kissing Congressman' Admits Mistake in Reelection Bid
- Clinton Rallies New England Women
- Ebola in New York, Mali Raises Travel Jitters
- Market Jolt Offers a Reality Check for Investors
- Tech Workers Needed in South Florida
- Steris to Add 100 Jobs in Birmingham