International Business Machines Corporation

Cheng-Chung Song

Yu-Cheng Hsu

Michael T Benhase

Lokesh M Gupta

William Griswold Sherman

Andrew Dale Walls

John Norbert McCauley

Paul Henri Muench

Stephen LaRoux Blinick

Roman Alexander Pletka

Nikolas Ioannou

Sasa Tomic

Radu Ioan Stoica

Wenjeng Ko

Ioannis Koltsidas

Karl Allen Nielsen

Joseph S Hyde, Ii

Lawrence Yiumchee Chiu

Clement Lambert Dickey

Lucien Mirabeau

Daniel Anson Heffley

Ricky Dean Rankin

Matthew J Kalos

Roger Gregory Hathorn

Timothy J Fisher

Charles John Camp

Richard Anthony Ripberger

Robert Akira Kubo

Zhi Jun Wang

Xue Feng Gao

Chao Li

Kurt Allen Lovrien

Timothy Roy Block

Gary William Batchelor

Louis Alonso Rasor

David Ray Kahler

Chiahong Chen

Carol Spanel

Michael John Jones

Juan J Ruiz

William Henry Travis

Sean Patrick Riley

Chao G Li

Yao Peng

Radha K Ramachandran

Ronald James Chapman

Wen-jeng Ko

Herve Philippe Andre

Robert W Tillerson

Techniques for selecting a storage node of a storage system to store data include applying a first function to at least some data chunks of an extent to provide respective first values for each of the at least some data chunks. A storage node, included within multiple storage nodes of a storage system, is selected to store the extent based on a majority vote derived from the respective first values.

Techniques for improving deduplication efficiency in a storage system with multiple storage nodes

A controller of a data storage system generates fingerprints of data blocks written to the data storage system. The controller maintains, in a data structure, respective state information for each of a plurality of data blocks. The state information for each data block can be independently set to indicate any of a plurality of states, including at least one deduplication state and at least one non-deduplication state. At allocation of a data block, the controller initializes the state information for the data block to a non-deduplication state and, thereafter, in response to detection of a write of duplicate of the data block to the data storage system, transitions the state information for the data block to a deduplication state. The controller selectively performs data deduplication for data blocks written to the data storage system based on the state information in the data structure and by reference to the fingerprints.

Workload optimized data deduplication using ghost fingerprints

Provided are a computer program product, system, and method for managing caching of extents of tracks in a first cache, second cache and storage device. A determination is made of an eligible track in a first cache eligible for demotion to a second cache, wherein the tracks are stored in extents configured in a storage device, wherein each extent is comprised of a plurality of tracks. A determination is made of an extent including the eligible track and whether second cache caching for the determined extent is enabled or disabled. The eligible track is demoted from the first cache to the second cache in response to determining that the second cache caching for the determined extent is enabled. Selection is made not to demote the eligible track in response to determining that the second cache caching for the determined extent is disabled.

Managing caching of extents of tracks in a first cache, second cache and storage

A data storage system includes a pre-cache and a plurality of storage devices across which a data storage array is distributed. In response to receipt of a write request specifying a logical address and write data, the data storage system buffers the write request among a plurality of write requests in the pre-cache without provisioning in the data storage array a physical extent corresponding to the logical address. A management node analyzes the plurality of write requests buffered in the pre-cache. In response to the analyzing identifying a first pattern of write requests, the management node provisions, in the data storage array, a first physical extent having a smaller grain size and destages the write data to the first physical extent. In response to the analyzing identifying a second pattern of write requests, the management node provisions a second physical extent having a larger grain size and destages the write data from the pre-cache to the second physical extent.

Optimizing thin provisioning in a data storage system through selective use of multiple grain sizes

In at least one embodiment, a data storage system has a plurality of non-volatile storage devices, a higher level controller and one or more lower level controllers. In response to one or more inputs, the higher level controller of the data storage system issues an initialization command specifying initialization of a storage extent. In response to the initialization command, one or more lower level controllers issue dataless initialization commands to the plurality of non-volatile storage devices. In response to the dataless initialization commands, multiple first non-volatile storage devices among the plurality of non-volatile storage devices each initialize a respective data portion of the storage extent and a second non-volatile storage device among the plurality of non-volatile storage devices initializes a data protection portion of the storage extent.

Efficient initialization of a thinly provisioned storage array

A mechanism is provided for performing de-duplication process on a set of non-volatile memories as part of another process routinely performed on the set of non-volatile memories. A hash value of data stored at a first physical location in a non-volatile memory in the set of non-volatile memories is received from a non-volatile memory controller associated with the non-volatile memory. Responsive to the hash value matching one or more existing hash values for data stored at one or more other physical locations in the set of non-volatile memories, an optimal physical location is identified from the first physical location and the one or more other physical locations. Responsive to identifying the optimal physical location, a set of logical addresses associated with the hash values is updated to point to the optimal physical location. The non-optimal physical locations are further invalidated in order that the non-optimal physical locations are erased.

De-duplication as part of other routinely performed processes

For data processing in a computing storage environment by a processor device, the environment incorporating at least high-speed and lower-speed caches, and managed tiered levels of storage, groups of data segments are migrated between the tiered levels of storage such that clumped uniformly hot ones of the groups of data segments are migrated to use a Solid State Drive (SSD) portion of the tiered levels of storage; uniformly hot groups of data segments are determined using a first, largest granulated, heat map for a selected one of the group of the data segments; a second heat map, which is smaller than the first and having the largest granularity of the first heat map, is used to determine the clumped hot groups; and sparsely hot groups are determined when neither the first heat map nor the second heat map are hotter than the first and second predetermined thresholds, respectively.

Tiered caching and migration in differing granularities

For data processing in a distributed computing storage environment by a processor device, the distributed computing environment incorporating at least high-speed and lower-speed caches, and managed tiered levels of storage, groups of data segments and clumped hot ones of the data segments are migrated between the tiered levels of storage such that uniformly hot ones of the groups of data segments are migrated to use a Solid State Drive (SSD) portion of the tiered levels of storage; uniformly hot groups of data segments are determined using a first, heat map for a selected one of the group of the data segments; and a second heat map is used to determine the clumped hot groups.

For data processing in a computing storage environment by a processor device, the computing storage environment incorporating at least high-speed and lower-speed caches, and tiered levels of storage, groups of data segments are migrated between the tiered levels of storage such that if a selected group is cached in the lower-speed cache and is determined to become uniformly hot, migrating the selected group from the lower-speed cache to the SSD portion while refraining from processing data retained in the lower-speed cache until the selected group is fully migrated to the SSD portion.

Use of differing granularity heat maps for caching and migration

Provided are techniques for receiving an error inject script that describes one or more error inject scenarios that define under which conditions at least one error inject is to be executed and compiling the error inject script to output an error inject data structure. While executing code that includes the error inject, an indication that an event has been triggered is received, conditions defined in the one or more error inject scenarios are evaluated using the error inject data structure, and, for each of the conditions that evaluates to true, one or more actions defined in the error inject script for the condition are performed.

Framework for a software error inject tool

For data processing in a computing storage environment by a processor device, the computing storage environment incorporating at least high-speed and lower-speed caches, and tiered levels of storage, groups of data segments are migrated between the tiered levels of storage such that uniformly hot ones of the groups of data segments are migrated to utilize a Solid State Drive (SSD) portion of the tiered levels of storage, while sparsely hot ones of the groups of data segments are migrated to utilize the lower-speed cache.

For data processing in a distributed computing storage environment by a processor device, the distributed computing environment incorporating at least high-speed and lower-speed caches, and managed tiered levels of storage, groups of data segments and clumped hot ones of the data segments are migrated between the tiered levels of storage such that uniformly hot ones of the groups of data segments are migrated to use a Solid State Drive (SSD) portion of the tiered levels of storage; uniformly hot groups of data segments are determined using a first, largest granulated, heat map for a selected one of the group of the data segments; and a second heat map, which is smaller than the first and having the largest granularity of the first heat map, is used to determine the clumped hot groups.

Growing community of inventors

Tucson, AZ, United States of America

Cheng-Chung Song