International Business Machines Corporation

Himabindu Pucha

Karan Gupta

Prasenjit Sarkar

Wendy Ann Belluomini

Joseph S Glider

Renu Tewari

Anna Povzner

Frank Bernhard Schmuck

Dean Hildebrand

Roger Lee Haskin

Reshu Jain

David Darden Chambliss

Rui Zhang

Dinesh Kumar Subhraveti

Prashant Pandey

Jorge Guerra Delgado

Rajagopal Ananthanarayanan

Mansi Ajit Shah

Amar Phanishayee

Embodiments relate to supporting transaction data committed to a stable storage. Committed data in the cluster is stored in the persistent cache layer and replicated and stored in the cache layer of one or more secondary nodes. One copy is designated as a master copy and all other copies are designated as replica, with an exclusive write lock assigned to the master and a shared write lock extended to the replica. An acknowledgement of receiving the data is communicated following confirmation that the data has been replicated to each node designated to receive the replica. Managers and a director are provided to support management of the master copy and the replicas within the file system, including invalidation of replicas, fault tolerance associated with failure of a node holding a master copy, recovery from a failed node, recovered of the file system from a power failure, and transferring master and replica copies within the file system.

Write performance in fault-tolerant clustered storage systems

A method for managing storage allocation includes adaptively determining, by a storage device processor, a region width across disk spaces for a group of storage devices that is inversely proportional to a number of nodes sharing a particular storage device. An adaptive storage device allocation region of the particular storage device is created based on the determined region width across the disk spaces for the group of storage devices.

Generalized storage allocation for multiple architectures

Embodiments of the invention relate to supporting transaction data committed to a stable storage. Committed data in the cluster is stored in the persistent cache layer and replicated and stored in the cache layer of one or more secondary nodes. One copy is designated as a master copy and all other copies are designated as replica, with an exclusive write lock assigned to the master and a shared write lock extended to the replica. An acknowledgement of receiving the data is communicated following confirmation that the data has been replicated to each node designated to receive the replica. Managers and a director are provided to support management of the master copy and the replicas within the file system, including invalidation of replicas, fault tolerance associated with failure of a node holding a master copy, recovery from a failed node, recovered of the file system from a power failure, and transferring master and replica copies within the file system.

Embodiments of the invention relate to storage allocation in a storage system. One embodiment includes adaptively determining, by a storage device processor, a region height across disk spaces of a group of storage devices that is proportional to a number of connected storage devices. The storage device processor adaptively determines a region width across the disk spaces for the group of storage devices that is inversely proportional to a number of nodes sharing a particular storage device.

Maintaining high availability of objects for both read and write transactions. Secondary copies of cached objects are created and maintained on disks of a secondary caching node and in remote data storage. In response to an update request, the secondary copies of cached objects are updated. Secondary cached objects are synchronously invalidated in response to the update request, and the update is asynchronously propagated to a secondary caching node.

Reducing memory overhead of highly available, distributed, in-memory key-value caches

Embodiments relate to cluster-centric tiered storage with a flexible tier definition to support performance of transactions. Object data is distributed in a multi-tiered shared-nothing cluster. Hierarchical tiers of data storage are assigned different roles within the hierarchy. The tiers are arranged according to a number of cycles required to access a tier. The tiers are managed globally across the cluster and objects are placed in tiers according to a flexible tier definition and the tier arrangement. The probability of object access is computed for objects, and objects are placed on different tiers responsive to the computation and the number of cycles required to access the tier. Objects are moved between tiers responsive to a probability frequency of object access.

Enhancing analytics performance using distributed multi-tiering

Embodiments of the invention relate to storage allocation in a storage system. One embodiment includes generating a request for storage space allocation in a particular storage device by a first node. An owner node associated with the particular storage device is determined by a first allocation client associated with the first node. The request is sent by the first allocation client to a second allocation client associated with the owner node. A storage device allocation region of the particular storage device is created, the allocation region comprising a height proportional to storage devices the owner node and the second allocation client are coupled with, and a width that is inversely proportional to a number of nodes sharing the particular storage device.

Embodiments of the invention relate to cluster-centric tiered storage with a flexible tier definition to support performance of transactions. Object data is distributed in a multi-tiered shared-nothing cluster. Hierarchical tiers of data storage are assigned different roles within the hierarchy. The tiers are managed globally across the cluster and objects are placed in tiers according to a flexible tier definition. The probability of object access is computed for objects, and objects are moved to different tiers responsive to the computation to minimize system runtime. The location of an object is further optimized in response to an access request.

Embodiments of the invention relate to optimizing a file system for different types of applications in a compute cluster using dynamic block size granularity. An exemplary embodiment includes reserving a predetermined number of storage allocation regions for each node in a cluster, wherein each storage allocation region comprises a set of contiguous data blocks on a single storage disk of locally attached storage; using a contiguous set of data blocks on a single storage disk of locally attached storage as a file system's block allocation granularity for striping data to be stored in the file system for a compute operation in the cluster; and using a single data block of a shared storage subsystem as the file system's block allocation granularity for striping data to be stored in the file system for a data access operation in the cluster.

Optimizing a file system for different types of applications in a compute cluster using dynamic block size granularity

In one embodiment, a configuration adviser provides configuration options for a multi-tiered storage system that includes a number of different storage tiers, each of which include storage devices of a particular storage type. Data access information for storage extents to be stored in the storage system are received. Resource information for available storage tiers in the storage system to place the storage extents on are also received. A cost incurred by the storage system for placing each of the storage extents on each of the storage tiers is determined. Each storage extent is assigned to a particular one of the storage tiers that would incur the lowest cost to the storage system for storing the storage extent. For each storage tier, a minimum number of storage devices are selected, within the assigned storage tier, that would satisfy data access and capacity requirements for all storage extents assigned to that tier.

Multi-tier storage system configuration adviser

Embodiments of the invention relate to dynamic power management of storage volumes and disk arrays in a storage subsystem to mitigate loss of performance resulting from the power management. The volumes and arrays are prioritized, and in real-time power is selectively reduced in response to both the prioritization and an energy savings goal. A feedback loop is provided to dynamically measure associated power gain based upon a lowering of power consumption, and device selection may be adjusted based upon received feedback.

Method for on-demand energy savings by lowering power usage of at least one storage device in a multi-tiered storage system

On-demand storage system energy savings

Embodiments of the invention relate to block layout and block allocation in a file system to support transparency of application processing. At least one copy of an application is replicated in a write affinity region of a secondary server, and at least one copy of the application is replicated in a wide striping region across a cluster file system. When the application is subject to failure, application processing is transferred from the failure location to the write affinity copy. At the same time, the failed application is rebuilt using the wide striping replication of the application. Once the application is rebuilt, processing may return to the failed location employing the rebuilt application.

Growing community of inventors

San Jose, CA, United States of America

Himabindu Pucha