86 results about "Shingled magnetic recording" patented technology

A “write-squeeze-verify” method is used for verification of the data that has been written in the annular bands of a shingled magnetic recording disk drive. The writing of data along a track overwrites a portion of the previously written track and thus “squeezes” the data of the previously written track to thereby form a “shingled data track” (SDT). The data in each SDT is read back and verified by performing an error correction check using error correction bits associated with the data written in the SDT, or by comparing the readback data with the data stored in memory. If the data read back is not verified, a write error counter is incremented and a write error frequency is calculated. One or more attempts to write the data can be performed. If the data in the SDT cannot be verified after the attempted rewrite(s), then a “re-try fail” is reported.

A shingled magnetic recording disk drive with sector error correction code (ECC) has the disk recording surface divided into multiple zones. Each zone is assigned a sector-ECC strength, i.e., a unique number of ECC sectors associated with a block of data sectors. The zone in which data is to be written is determined from the time average of the position-error signal (PES), which is an indication of the track misregistration (TMR) and thus the current environmental condition to which the HDD is subjected.

Inter-track interference cancelation is disclosed, including: receiving an input sequence of samples associated with a track on magnetic storage; using a processor to generate inter-track interference (ITI) data associated with a first side track including by performing a correlation between the input sequence of samples and a sequence of data associated with the first side track.

One embodiment facilitates a write operation in a shingled magnetic recording device. During operation, the system receives, by the storage device, data to be written to the storage device and access-frequency information associated with the data, wherein the storage device includes a plurality of concentric tracks. The system distributes a plurality of spare sector pools among the plurality of concentric tracks. The system selects a track onto which to write the data based on the access-frequency information, wherein data with a highest access-frequency is written to an outer track. The system appends the data at a current write pointer location of the selected track, thereby facilitating an enhanced data placement for subsequent access in the storage device.

A hard disk drive that includes a disk with data written onto a plurality of tracks, a spindle motor that rotates the disk, and a head that is coupled to the disk. The disk drive also includes a circuit that writes data onto a first writable shingle band of tracks if the first writable shingle band is adjacent to a guard band of tracks. The first writable shingle band includes a number of tracks that is a function of a head width. The guard band of tracks is capable of becoming a writable shingle band. Changing the designation of a shingle band between guard and writable creates floating guard bands. The creation of floating guard bands allows for the writing of a single band without having to move and restore adjacent tracks until reaching a fixed guard band as required in the prior art.

Storage architecture for bit patterned media uses both erase band and shingled magnetic recording. A hard disk drive may comprise a disk having bit patterned media with a plurality of data tracks arrayed in architecture pages having at least one of erase band mode (EBM), shingled mode (SM) and unallocated space. An actuator has a head for writing data to the data tracks of the bit patterned media. A control system monitors, reallocates and reconfigures the architecture pages from EBM, SM or unallocated space to a different one of EBM, SM or unallocated space to enhance performance of the hard disk drive.

One embodiment facilitates a write operation in a shingled magnetic recording device. During operation, the system receives, by a controller module of the device, a request to write first data, wherein the device has a plurality of bands with overlapping tracks for storing data. In response to determining that the first data is updated data corresponding to original data stored in a first band, the system appends the updated data to a second band with available storage space. The system merges the updated data with the original data.

An indirection system in a shingled storage device is described that uses an efficient algorithm to map LBAs to DBAs based on a predetermined rule or assumption and then handles as exceptions LBAs that are not mapped according to the rule. The assumed rule is that a fixed-length set of sequential host LBAs are located at the start of an I-track. Embodiments of the invention use two tables to provide the mapping of LBAs to DBAs. The mapping assumed by the rule is embodied in the LBA Block Address Table (LBAT) which gives the corresponding I-track address for each LBA Block. The LBA exceptions are recorded using an Exception Pointer Table (EPT), which gives the pointer to the corresponding variable length Exception List for each LBA Block. The indexing into the LBAT and the EPT is made efficient by deriving the index from the LBA by a simple arithmetic operation.

To provide enhanced operation of data storage devices and systems, various systems, apparatuses, methods, and software are provided herein. In a first example, a data storage device is presented with storage media comprising a cache storage region and a shingled magnetic recording (SMR) storage region that is divided into burst zones. A storage control system receives write operations and accumulates write data in the cache storage region until a commit threshold condition. Responsively, the storage control system transfers the write data into a burst zone of the SMR storage region, and verifies the write data written into the burst zone once the burst transfer is complete. Responsive to data verify errors in the write data written into the burst zone, the storage control system writes data associated with the data verify errors into an exception location associated with the burst zone.

A shingled magnetic recording hard disk drive that uses writeable cache tracks in the inter-band gaps between the annular data bands minimizes the effect of far track erasure (FTE) in the boundary regions of annular data bands caused by writing to the cache tracks. Based on the relative FTE effect for all the tracks in a range of tracks of the cache track being written, a count increment (CI) table or a cumulative count increment (CCI) table is maintained. For every writing to a cache track, a count for each track in an adjacent boundary region, or a cumulative count for each adjacent boundary region, is increased. When the count value for a track, or the cumulative count for a boundary region, reaches a predetermined threshold the data is read from that band and rewritten to the same band.

A technique for recovering of “squeezed” sectors in a set of sequential sectors such as are used in Shingled Magnetic Recording (SMR) is described. Embodiments of the invention use a programmable erased sector recovery scheme, which is a concatenation of a “Cauchy-type” track erasure correction code, together with a media-error correction code that generates N-weighted parity-sectors per track and is capable of replacing up to N-erased sectors per track in any possible combination.

A method of encoding user data into a first set of codewords using a first code, generating a first set of parity information based at least in part on the first set of codewords and at least a second code, and writing at least parity information associated with the first set of parity information to shingled magnetic recording storage. A method of performing decoding on a first set of read-back signal data read back from shingled magnetic recording storage and associated with a first set of codewords, and if decoding of at least one read-back signal in the first set of read-back signal data fails, performing decoding on at least some of a second set of read-back signal data associated with a set of parity information.

The invention relates to a data storage method and device, which are used for solving the problem of multiple read-write amplification and performance reduction of a storage system when the shingled recording technology is combined with the LSM tree technology. The data storage method comprises the following steps: determining a target band capable of storing a target sorted string table SSTable group from a shingled magnetic recording SMR disk, wherein the target SSTable group is stored in a log structured LSM tree, the LSM tree comprises at least two layers, each layer comprising at least one SSTable, and at least two SSTables in each layer of which the key value range is within the key value range of one SSTable in the upper layer are taken as one SSTable group; and storing the target SSTable group in the target band.