International Business Machines Corporation

Steven Ross Bentley

Keisuke Tanaka

Roy Daron Cideciyan

Simeon Furrer

Pamela Ruth Nylander-Hill

Sushama M Paranjape

Robert Allan Hutchins

Stephen C West

Fernando Quintana

Kenji Ohtani

Charles E Bailey

Rickey W Murray

Pantas Sutardja

Daniel James Winarski

Nhan Xuan Bui

Allen Keith Bates

Kazuhiro Tsuruta

Eiji Ogura

James Mitchell Karp

Reed Alan Hancock

Robert Allen Hutchins

Hirokazu Nakayama

Paul Merrill Greco

Alex Chliwnyj

Paul Joseph Seger

Constantin Michael Melas

Mitsuhiro Nishida

Hisato Matsuo

Scott Jeffrey Schaffer

Steven Douglas Johnson

Katsuhiko Hagiwara

Shuhei Nadehara

Fahnmusa Christian Jangaba

Ronald K Rhodes

Christine Renee Knibloe

Gary W Callsen

Demura Masayuki

David M Fickle

Randall Scot Doan

In one embodiment, a system includes a controller and logic integrated with and/or executable by the controller. The logic is configured to perform iterative decoding on encoded data to obtain decoded data. The logic is also configured to perform post-decoding error diagnostics on a first portion of the decoded data in response to not obtaining a valid product codeword in the first portion after the iterative decoding of the encoded data. Other systems, methods, and computer program products for producing post-decoding error signatures are presented in accordance with more embodiments.

Post-decoding error check with diagnostics for product codes

In one embodiment, an apparatus includes a controller and logic integrated with and/or executable by the controller. The logic is configured to perform track-dependent erasure decoding on encoded data based on detection of one or more time-varying signal quality issues associated with at least one of a plurality of tracks read simultaneously from a data storage medium. In another embodiment, a method includes determining, by a magnetic tape drive, track signal quality reliability for a plurality of tracks read simultaneously from a magnetic tape medium. In addition, the method includes performing, by the magnetic tape drive, track-dependent erasure decoding on encoded data based on detection of one or more time-varying signal quality issues associated with at least one of the plurality of simultaneously-read tracks.

Track-dependent decoding

Various embodiments for data error recovery in a tape storage system, by a processor device, are provided. In one embodiment, a method comprises, in a tape storage system using an iterative hardware and microcode decoder, initializing at least one iterative decode cycle on the buffered dataset when an uncorrectable read error occurs; and for a next iterative decode cycle, building upon data corrections obtained in a previous iterative decode cycle.

Iterative decode and post-decode microcode cut and paste optimization

In one embodiment, a method includes reading, using a plurality of read sensors of a magnetic head in a read channel of a tape drive, encoded data from a plurality of tracks of a magnetic tape medium simultaneously. Also, the method includes monitoring statistics for the read channel, the statistics monitored for the read channel including exponentially averaged mean squared error (MSE) or signal-to-noise ratio (SNR) for each simultaneously read track. Moreover, the method includes performing track-dependent erasure decoding on the encoded data based on detection of one or more time-varying signal quality issues associated with at least one of the plurality of tracks read simultaneously from the magnetic tape medium. The one or more time-varying signal quality issues are determined based on side information about reliability of detected bytes within C1 codewords of each simultaneously-read track.

Various embodiments for data error recovery in a tape storage system, by a processor device, are provided. In one embodiment, a method comprises modifying erasure control configuration settings upon rereading a buffered dataset having passed through at least one microcode-initiated iterative decode cycle. X microcode-initiated iterative decode cycles are initiated on the buffered dataset while the tape is stopped, where x comprises at least one of a plurality of the microcode-initiated iterative decode cycles. The x microcode-initiated decode cycles are initiated on the buffered dataset until a predetermined error correction threshold is reached.

Microcode data recovery strategies for use of iterative decode

Various embodiments for data error recovery in a tape storage system, by a processor device, are provided. In one embodiment, a method comprises, in a tape storage system using an iterative hardware decoder, dynamically initializing at least one microcode-initiated decode cycle on a buffered dataset, each iterative decode cycle providing an error feedback loop.

In one embodiment, a computer-implemented method includes writing a data set to a first write section of a magnetic medium and rewriting at least some of the data set as rewritten CWI-sets to a rewrite section of the magnetic medium. The data set includes a plurality of sub data sets, each sub data set including a data array organized in rows and columns. Each row of the data array includes four interleaved C1 codewords (a CWI-). A first portion of the data set is stored as CWI-sets to the first write section of the magnetic medium with first headers. Each rewritten CWI-set is stored to the rewrite section of the magnetic medium as a number of rewritten CWI-s having corresponding rewrite headers. Also, a length of any one of the rewrite headers is greater than a length of any one of the first headers.

Efficient multichannel data format using variable-length headers

In one embodiment, a computer program product includes a computer readable storage medium having program instructions embodied therewith. The computer readable storage medium is not a transitory signal per se. The program instructions are executable by a tape drive to cause the tape drive to perform a method. The method includes writing data to a first write section of a magnetic tape, at least some of the data being written in association with first headers. The method also includes selecting some of the data for rewrite based on detected errors. Moreover, the method includes rewriting the selected data to a rewrite section of the magnetic tape, the rewritten data being written in association with rewrite headers. A length of each of the rewrite headers is greater than a length of each of the first headers.

Various embodiments for data error recovery in a tape storage system, by a processor device, are provided. In one embodiment, a method comprises, in a tape storage system using a hardware-initiated and microcode-initiated iterative decoder, initializing a microcode cut/paste algorithm on a buffered dataset by copying the buffered dataset to a reserved buffer segment; wherein the buffered dataset is a most corrected buffered dataset developed by at least one iterative decode cycle; and the buffered dataset, updated C1 data row error flags, and a correction status recovered by the at least one iterative decode cycle are cut and pasted over a working copy of a dataset resident in the reserved buffer segment.

Various embodiments for data error recovery in a tape storage system, by a processor device, are provided. In one embodiment, a method comprises, in a tape storage system using an iterative hardware decoder and an iterative microcode decoder, modifying erasure control configuration settings upon rereading a buffered dataset having passed through at least one microcode-initiated iterative decode cycle.

In one embodiment, a method includes loading first data into a first buffer of an interposer during a first time period and loading second data into a second buffer of the interposer and performing a first decoding operation on the first data using a first decoder during a second time period. The method includes loading third data into a third buffer of the interposer, performing the first decoding operation on the second data using the first decoder, and performing a second decoding operation on the first data using a second decoder during a third time period. Moreover, the method includes loading fourth data into a fourth buffer of the interposer, performing the first decoding operation on the third data using the first decoder, and performing the second decoding operation on the second data during a fourth time period. The first and second decoding operations are C1 or C2 decoding operations.

Pipelined architecture for iterative decoding of product codes

In one embodiment, a tape drive includes a magnetic head having a plurality of read sensors, each read sensor being configured to read data simultaneously. The tape drive also includes a controller and logic integrated with and/or executable by the controller. The logic is configured to receive encoded data read from a plurality of tracks of a magnetic tape medium simultaneously. The logic is also configured to perform priority-based decoding on the encoded data based on erasure coefficients associated with at least one codeword of the encoded data. In another embodiment, a controller-implemented method includes receiving encoded data read from a plurality of tracks of a magnetic tape medium simultaneously and performing priority-based decoding on the encoded data based on erasure coefficients associated with at least one codeword of the encoded data.

Priority-based decoding

In one embodiment, a system includes a controller and logic integrated with and/or executable by the controller. The logic is configured to perform iterative decoding on encoded data to obtain decoded data. At least three decoding operations are performed in the iterative decoding, with the decoding operations being selected from a group consisting of: C1 decoding and C2 decoding. The logic is also configured to perform post-decoding error diagnostics on a first portion of the decoded data in response to not obtaining a valid product codeword in the first portion after the iterative decoding of the encoded data. Other systems, methods, and computer program products for producing post-decoding error signatures are presented in accordance with more embodiments.

In one embodiment, a tape drive includes a magnetic head having a plurality of read sensors, each read sensor being configured to read data simultaneously, a controller, and logic integrated with and/or executable by the controller. The logic is configured to read, using the plurality of read sensors, encoded data from a plurality of tracks of a magnetic tape medium simultaneously. The logic is also configured to perform track-dependent erasure decoding on the encoded data based on detection of one or more time-varying signal quality issues associated with at least one of the plurality of tracks read simultaneously from the magnetic tape medium.

In one embodiment, a system includes a controller and logic integrated with and/or executable by the controller, the logic being configured to write a data set to a first write section of a magnetic medium, the data set including a plurality of sub data sets, each including a data array organized in rows and columns, each row of the data array including a CWI-4. A first portion of the data set is stored as CWI-4 sets to the first write section of the magnetic medium with first headers. The logic is also configured to rewrite at least some of the data set as rewritten CWI-4 sets to a rewrite section of the magnetic medium as rewritten CWI-4s having corresponding rewrite headers. A length of any one of the rewrite headers is greater than a length of any one of the first headers.

Growing community of inventors

Tucson, AZ, United States of America

Steven Ross Bentley