International Business Machines Corporation

Paul Hodges

Robert Wesley Shomler

John Scott Eggenberger

Mien Shih

Michael Garwood Hurley

Norman Kenneth Ouchi

Michael T Benhase

Brent Cameron Beardsley

Susan Kay Candelaria

Michael Howard Hartung

Bernd Lamberts

Arvind M Patel

Norman Ken Ouchi

William H Tetzlaff

Robert Baird

Gerald P Bozman

Richard Roland Guyette

James J Myers

Alexander S Lett

Joel H Cord

Werner J Schaeuble

David A Plomgren

Paul L Leung

Bruce M Henry

Thomas Beretvas

John S Eggenberger

Paul L Shaffer

High level commands for a disk drive are processed by a task manager program that parses them into low level subcommands (e.g., SCSI commands). The low level subcommands are. then presented to the command execution means of the disk drive for execution. The parsing or sub tasking of a high level command into a number of low level commands permits the efficient handling of subcommands that must be deferred until other subcommands are executed and of third party XOR operations.

Task control for high level commands in disk drives

A disk drive that is enabled by information contained in a mode page and in a command to access data blocks of either a small size (e.g., 512 bytes) or of a large size (e.g., 4,096 bytes). This allows legacy application software written for 512 byte sized blocks to run without change. By turning on the 4,096 byte sized block flag in the command, new software written to obtain improved performance benefits may also be used with the disk drive. The dual block capability is implemented in software.

Disk drive and method of multi-block size addressing

A method and apparatus for serializing access to disk arrays shareable among a plurality of RAID control units at a substantial reduction in intercontrol unit communication by (a) defining a lock function over the parity image blocks at each of the disk drives of a shared disk array; and (b) executing a path expression at each accessing control unit, the path expression includes requesting a lock from the drive on the parity image and enforcing a busy-wait until a lock is granted, executing the RAID function, and then releasing the lock.

Device level coordination of access operations among multiple raid control units

A data processing system having multiple independent paths for communication between a host and a plurality of storage devices where each path has its own queue for servicing requests generated by the host for accessing the storage devices. Each request is assigned a unique sequential ID before it is stored, along with its unique ID, in all the queues. Each storage device has a 'mailbox' register where the ID and the status of the latest request being carried out is stored. Queues are serviced and their status updated based on the content of the mailbox in each storage device. The combination of assigning a unique task ID to each request and a 'mailbox' register in each storage device allows the queue in each path to be completely out of sync with each of the queues in the other paths without causing data integrity problems, duplication of requests at the device level, or a need for complex locking schemes to keep the queues in sync with each other.

Multipath I/O storage systems with multiipath I/O request mechanisms

A system and method are provided for coordinating command for updating and reconstructing data in an array of storage devices connected to a computer system when multiple initiators issue commands. Sets of data blocks and corresponding parity blocks are arranged on a plurality of storage devices, such that a data block can be reconstructed using a set of data blocks and at least one parity block. When a new data block is written to the disk drive, a corresponding old parity block is write updated with an updated parity block. When reconstructing an unavailable data block an initiator issues a read command to each storage device having a data block or a parity block used for reconstructing the unavailable data block. Each read command is enqueued on a command queue for each storage device in a priority order coordinating each read command with any write update command for the same block so that a new data block and an old parity block are not read as part of the same parity stripe and vice versa.

System and method for reconstructing data in a storage array system

A data processing system incorporating the invention hereof, includes a host processor or server, connected to an array of disk drives which are further interconnected via a network. Each disk drive is logically divided into plural logical memory segments and, in the preferred embodiment, corresponding logical memory segments in each disk drive are arranged as a parity set in a RAID fashion. The host processor includes a driver which generates read and write commands to the array of disk drives. The commands include addresses which are in accord with ranges of addresses assigned to logical memory segments in each disk drive. Each disk drive stores an array configuration page which includes data that identifies a logical memory segment size; a first identifier for the disk drive that incorporates the respective memory segment; and a second identifier for the last disk drive which stores parity for the parity set including the logical memory segment. Each disk drive includes a controller for determining from the logical memory segment size and the second identifier, which logical memory segment in the respective memory stores parity data. The controller responds to an address accompanying a read or write command by mapping the address to a logical memory segment, other than the logical memory segment in the respective memory which stores parity data. Further, the controller, upon performing a write action to a logical memory segment, communicates directly with the disk drive that stores parity for the parity set of which the logical memory segment is a part and causes an update of that parity segment in accordance with the write action.

Raid-configured disk drive array wherein array control occurs at the

Multipath I/O storage systems with multipath I/O request mechanisms

A data storage system including a controller and a string of multiple DASDs, where each DASD is responsive to commands from the controller to perform local operations such as XOR functions and transferring data to other DASDs. A bus, which may comprise a serial or a parallel bus, electrically interconnects the controller and the DASDs. Each DASD includes a storage apparatus, an interface, a buffer, and processor. The buffer selectively receives data from the interface and the storage apparatus. In addition to other operations, the processor is responsive to commands from the controller to the buffer interface to perform an XOR operation on selected items of data in the buffer. The processor may additionally direct the results of the XOR operation to another DASD or to the controller. The data storage system of the invention may advantageously be operated to perform a task requiring the performance of multiple XOR operations, wherein each DASD performs an XOR operation on data and transmits the result to another DASD in accordance with a 'daisy chain' specified for that task. Hence, controller workload is reduced and computational efficiency is increased by distributing the XOR operations among the DASDs.

Data storage system with localized XOR function

A method for dynamically detecting loss of map integrity in a form of system-managed storage (SMS). In SMS, maps are used to define access paths to data and to allocate and reallocate storage resources among applications running thereon. The method steps include incorporating as an indivisible part of an overwriting commmand the duplication of map information by appending a portion of it to each data block in store, and detecting loss of map integrity as a function of a comparison mismatch between the portion stored with a counterpart data block and the map upon each read/write access.

Method for verification and restoration of directories in CPU system

A data processing system and method providing error protection for data transmitted between a processor and a buffer in one data format and transmitted between the buffer and a user device in a different data format. An adaptor is interposed between the processor and the buffer for transmitting to the buffer (i) successive data segments in the one data format, each ending with appended check bytes in a preselected cyclic redundancy code (CRC); and (ii) check bytes using the same CRC appended at the end of each segment in the different data format to create in the buffer records which are a composite of both formats, but viewed as in the one data format by the processor and as in the different data format by the user device. The boundaries of the segments in each format must be known to the adaptor. Since both formats use the same CRC, CRC bytes for each segment in each data format will provide an identical preselected value in the absence of a detectable error. A second adaptor, which does not and need not know the boundaries of the data segments in the one data format, is interposed between the buffer and the user device for transmitting the data between the buffer and user device in the different data format.

Common error protection code for data stored as a composite of different

Often in DASD subsystems, circumstances can occur which prevent the full transfer of the required data from channel to subsystem during a write operation. The disclosed methods prevents this data from later being read and treated as valid data by the host processor. This is achieved by marking data as invalid within the storage subsystem once it is determined that a channel error has occured. Subsequently, upon reading that data, the host processor can be made aware of the data invalidity and treat the data accordingly. In a second embodiment, invalid data is discarded rather than stored over the previous valid version of the data, before it is ever stored on DASD.

Invalid data detection, recording and nullification

A method is disclosed for correcting multibyte errors in a magnetic medium on which data is recorded in variable length blocks that comprise subblocks of data bytes and corresponding check bytes and include error correction code (ECC) for which ECC syndromes are generated during reading. A sequence of N sequential parity check bytes is written at the end of each block. After ECC syndromes are generated during reading, parity syndromes are generated by comparing parity check bytes computed from data bytes and check bytes as read with the parity check bytes as written. When a long-burst error occurs, a pointer points to the first of the N consecutive bytes in a block that could have been influenced by the error burst. After correcting correctable errors in all subblocks not affected by the N bytes identified by the pointer, and adjusting the parity syndromes for errors thus corrected, the adjusted parity syndromes are used to correct the errors in the N bytes indicated by the pointer. Unused ECC syndromes are then adjusted for errors corrected by the adjusted parity syndromes and used to correct all correctable errors then remaining.

Method and system for correcting long bursts of consecutive errors

A system is disclosed for generating a plurality of error correcting check ECC bytes from a block of data presented to the system in serial by byte form. The system employs a plurality of ECC channels which operate in parallel with the channels generating check bytes from interleaved subsets of the data block. One channel generates an ECC parity check byte for each interleaved subset while another channel generates an ECC locator check byte for each interleaved subset of data. The ECC locator check byte for each subset represents the parity or modulo 2 sum of bit positions which are selected systematically in accordance with a predefined m sequence which is unique to each channel that generates locator check bytes. Error patterns greater than the number of bits in one byte are correctable, as are error patterns which are less than the number of bits in one byte but extend across byte boundaries of two adjacent bytes in different subsets.

Error correcting system for serial by byte data

Apparatus for converting binary digital data from one form to another according to a variable word length code of fixed rate comprises word position indicating circuitry for indicating predetermined word position states of data in a data storage unit and conversion circuitry for sequentially converting a constant number of input bits at a time into the corresponding number of bits in said other form in dependence upon the contents of the data storage unit and the word position indication. The constant number of bits is less than the number of bits in the longest variable length word in the input data code form. The word position indicating circuitry either generates an updated function representing the current position of the word boundary in the data storage unit or recognizes the boundary from distinctive word ending patterns in the data storage unit.

Growing community of inventors

San Jose, CA, United States of America

Paul Hodges