International Business Machines Corporation

Florent Rostagni

John P Wilkinson

Ben Sasson

Andrea Sipka

Carlos Francisco Fuente

Lee Jason Sanders

Paul Nicholas Cashman

Gordon Douglas Hutchison

Christopher B E Beeken

Joanna Katharine Brown

Gareth John Nicholls

Dominic Tomkins

Alexander Henry Ainscow

Roderick Guy Charles Moore

Mark Keith Elliott

Leonor Maria Quintais

A method, computer program product, and computer system for copying of data in a storage system between storage drives that have embedded compression. The method includes determining that a copy operation of data is required from a source drive to a target drive and determining that the source drive and the target drive use the same compression algorithm. The method includes: reading the data in a compressed form from the source drive including bypassing a decompression algorithm; transmitting the compressed data from the source drive to the target drive; and receiving and writing the compressed data at the target drive including bypassing a compression algorithm at the target drive.

Copying of data in a storage system between storage drives that have embedded compression

Embodiments of the present invention provide concepts for handling a handover of ownership of data from a source to a referrer in a data deduplication environment. By performing a handover of the ownership of the data from the source to the referrer, the number of processes required to access the data may be reduced and so the performance of the system may be improved. The identification of a source for performing the handover on may be performed by way of a volatile cache.

Data deduplication

An apparatus for dynamically adjusting a journal snapshot window is disclosed. A computer-implemented method and computer program product also perform the functions of the apparatus. According to an embodiment of the present invention, a failure detection module detects a potential failure of at least a portion of a non-volatile storage device. The non-volatile storage device includes a journal for tracking metadata changes for data that is stored on the non-volatile storage device prior to committing the metadata changes to metadata for the non-volatile storage device. A notification receiving module receives an indication of the potential failure of the non-volatile storage device. A snapshot module adjusts a snapshot window for the journal in response to the indication of the potential failure. The snapshot window includes a set of metadata changes in the journal that have not been committed to the metadata of the non-volatile storage device.

Dynamically adjusting a journal snapshot window

Method and system are provided for data validation during data recovery in a log-structure array (LSA) storage system. The method includes reading a log record of a recovery log for a logical address to obtain a physical address at a storage backend for the data at the time of the log record; reading reference metadata at the obtained physical address, wherein the reference metadata indicates the logical address that last wrote data to the physical address; validating that the physical address for the log record contains valid data for the logical address of the log record by comparing the logical address of the reference metadata to the logical address of the log record; and replaying the log record if the physical address is validated for recovery of virtual domain logical metadata mapping the logical address to the physical addresses of the log record.

Data validation during data recovery in a log-structured array storage system

Method and system are provided for storage handling guidance for host input/output (I/O) operations. The method includes: providing a guidance array of indications of storage operations handling instructions, the guidance array having multiple dimensions of performance characteristics with each dimension having multiple levels; and associating a reference vector with one or more I/O operations, wherein the reference vector points to a level for each dimension of the array to obtain an indication of a storage operations handling instruction at an intercept of the dimension levels for application by a storage system controller for the one or more I/O operations.

Storage handling guidance for host input/output operations

Embodiments of the present invention disclose a method, a computer program product, and a computer system for data synchronization. A first data storage device reads a first data region and generates a first hash of the first data region before transmitting the first hash to a second data storage device. The second data storage device reads a second data region corresponding to the first data region and generates a second hash of the second data region. The second data storage device then determines whether the first hash matches the second hash and, based on determining that the first hash does not match the second hash, transmits data of the second data region to the first data storage device. The first data storage device applies the data of the second data region, thereby synchronizing the first data storage device and the second data storage device.

Synchronization of block based volumes

A computer-implemented method of storing data on a storage device can receive an I/O request to read data from or write data to a data storage device. The data can be stored in special sectors as allocated sectors and regular sectors as a log-structured array (LSA). If the I/O request is to read data from the storage device, a determination can be made as to whether the data to be read is stored as a special sector, which can be read from the LSA, or as a regular sector, which can be can be read from the allocated sectors. If the I/O request is to write data to the storage device, a determination can be made as to whether the data to be written is stored as a special sector, to the LSA or as a regular sector, to the allocated sectors.

Hybrid log-structured array and allocated storage device

A multi-node data storage system comprises a first data storage system having an owner node and a backup node in a first location coupled to a second data storage system having an owner node and a backup node in a second location. Each storage system includes a copy of the same data volume. A failure of a node of the multi-node storage system is detected. An outstanding write request to the first storage system is identified. If the owner node in the first storage system fails, it is determined whether the outstanding write corresponds to a host write to the backup node of the first storage system. If so, a retransmission message is sent to the second storage system. Otherwise, the data region associated with the outstanding write request is read from the first storage system, and a resynchronization message is sent to the second storage system.

Handling node failure in multi-node data storage systems

A process compares respectively corresponding data blocks in redundant data storage systems having ongoing I/O activity, to detect discrepancies between blocks that are supposed to be identical. Storage systems have replica copies of the data. A hashing algorithm processes a first data block from to produce a first hash. The hashing algorithm processes the second data block (a replica of the first data block) to produce a second hash. The first and second hashes are compared to determine whether or not the underlying data blocks match each other. If a write command directed to the data blocks is received during the overall comparison process, the comparison process is halted to permit the write command to proceed without adverse impact from the comparison process. The comparison process is then restarted.

Comparison of block based volumes with ongoing inputs and outputs

Storage sites are allocated pairs of nodes in which one node of the pair acts as an owner node while the other a backup node. When a local owner node receives a colliding write, the local owner node obtains a lock on the modified database and transfers both the write data and metadata to a remote owner node. The remote owner node returns a write complete message and the local site owner unlocks the modified database. When a local backup node receives a colliding write, the local backup node requests a lock from the local owner node and sends the write data to the remote owner node while the local owner node sends the write metadata to the remote owner node. The remote owner node then returns a write complete message to the local backup node which then requests the modified database be unlocked by the local owner node.

Growing community of inventors

Eastleigh, United Kingdom

Florent Rostagni