Abstract: A computer-implemented method according to one embodiment includes identifying a request to migrate data associated with a volume from a source storage pool that has a first rank extent size to a destination storage pool that has a second rank extent size greater than the first rank extent size. Additionally, the method includes creating a correspondence between logical volume extents of the volume and physical offset locations within rank extents of the destination storage pool. Further, the method includes migrating data from one or more ranks of the source storage pool to one or more ranks of the destination storage pool, according to the correspondence between the logical volume extents of the volume and the physical offset locations within the rank extents of the destination storage pool.

Abstract: A computer-implemented method, according to one approach, includes: identifying a request to convert an existing storage pool having a smaller first rank extent size into an intended storage pool having a larger second rank extent size. Volume extents in the existing storage pool are reordered such that physical rank extents of the volumes align with boundaries between the rank extents of the intended storage pool. Each of the ranks in the existing storage pool are converted from the first to the second rank extent size. Thereafter, for each of the volumes: a temporary pause/lock is performed on a VST entry which corresponds to the given volume, and an existing structure associated with the VST entry which corresponds to the first rank extent size is removed. Furthermore, the VST entry is populated with a new structure which corresponds to the second rank extent size.

Abstract: One general aspect of device reservation state preservation in accordance with the present description, provides for an intermediate reservation state, referred to herein as a “peer” reservation state, which may be maintained by a storage controller in the event of a total loss of communication connectivity to the reserving host so long as a peer or partner storage controller of a mirror relationship still has communication connectivity to the host. The peer reservation state as used herein, is a reservation state intermediate between a full reservation state for a device, and a fully released state in which a reservation of the device has been completely released. Other features and aspects may be realized, depending upon the particular application.

Abstract: A method for converting a redundant array of independent disks (RAID) to a more robust RAID level is disclosed. Such a method identifies, in a data storage environment, higher risk storage drives having a failure risk above a first threshold. The method determines a number of the higher risk storage drives that are contained within a RAID array of the data storage environment. The method determines whether the number exceeds a second threshold. The method also determines whether a destage rate associated with the RAID array is below a third threshold. In the event the number exceeds the second threshold and the destage rate is below the third threshold, the method converts the RAID array to a more robust RAID level. A corresponding system and computer program product are also disclosed.

Abstract: A method for reducing incidents of data loss in RAIDs of differing RAID levels is disclosed. Such a method identifies, in a data storage environment, first and second sets of RAIDs. The first set contains RAIDs providing more robust data protection (e.g., RAID-6 arrays), and the second set contains RAIDs providing less robust data protection (e.g., RAID-5 arrays). The method identifies, in the data storage environment, higher risk storage drives having a failure risk above a threshold and lower risk storage drives having a failure risk below the threshold. The method swaps higher risk storage drives from RAIDs of the second set with lower risk storage drives from RAIDs of the first set. In certain embodiments, the swapping may be performed such that no RAID of the second set includes more than a selected number of higher risk storage drives. A corresponding system and computer program product are also disclosed.

Abstract: A method for reducing incidents of data loss in redundant arrays of independent disks (RAIDs) having the same RAID level is disclosed. In one embodiment, such a method identifies, in a data storage environment, a set of RAIDs having a common RAID level. The method also identifies, in the set of RAIDs, higher risk storage drives having a failure risk above a threshold and lower risk storage drives having a failure risk below the threshold. The method swaps, within the RAIDs, higher risk storage drives with lower risk storage drives to more evenly distribute higher risk storage drives across the RAIDs. A corresponding system and computer program product are also disclosed.

Abstract: A method for reporting incidents of data loss in a storage environment comprising redundant arrays of independent disks (RAIDs) is disclosed. In one embodiment, such a method monitors storage drive failures in a storage environment. For a storage drive failure detected in the storage environment, the method reports the RAID type in which the storage drive failure occurred and whether data loss occurred in the RAID as a result of the storage drive failure. In certain embodiments, the method reports whether the data loss could have been prevented had the RAID type been converted to a more robust RAID type. In other or the same embodiments, the method reports whether the data loss was prevented by the RAID type. A corresponding system and computer program product are also disclosed.

Abstract: In response to an occurrence of a failure in a storage controller, an input on a plurality of attributes of the storage controller at a time of occurrence of the failure is provided to a machine learning module. In response to receiving the input, the machine learning module generates a plurality of output values corresponding to a plurality of recovery mechanisms to recover from the failure in the storage controller. A recovery is made from the failure in the storage controller, by applying a recovery mechanism whose output value is greatest among the plurality of output values that are generated by the machine learning module.

Abstract: A plurality of tracks that are to be copied to a backup volume are aggregated in a container data structure. The plurality of tracks are stored physically contiguously in a single Redundant Array of Independent Disks (RAID) stride. Mapping metadata is updated in the backup volume to indicate how logical tracks of the backup volume correspond to physical tracks stored in the RAID stride.

Abstract: A computer-implemented method for implementing a RAID array, according to one approach, includes: predicting a compression ratio of data to be stored in the RAID array, and using the predicted compression ratio and a physical storage capacity of the RAID array to calculate a maximum virtual storage capacity of the RAID array. The maximum virtual storage capacity is used to establish an effective storage capacity of the RAID array. One or more instructions to store compressed data in the RAID array are also sent. In response to occurrence of a predetermined event, a current compression ratio of the compressed data stored in the RAID array and the physical storage capacity of the RAID array are used to calculate a current virtual storage capacity of the RAID array. Furthermore, the effective storage capacity of the RAID array is scaled based on the current virtual storage capacity.

Abstract: A computer-implemented method according to one embodiment includes identifying a request to migrate data associated with a volume from a source storage pool having a first rank extent size to a destination storage pool having a second rank extent size smaller than the first rank extent size, creating a correspondence between logical volume extents of the volume and physical offset locations within rank extents of the source storage pool, and migrating data from one or more ranks of the source storage pool to one or more ranks of the destination storage pool, utilizing the correspondence between the logical volume extents of the volume and the physical offset locations within the rank extents of the source storage pool.

Abstract: Provided are a computer program product, system, and method for reducing a rate at which requests are sent from one processor to another to reduce processor utilization at the processor receiving the requests. A determination is made as to whether processor utilization at the first processor exceeds a utilization threshold. If so, a determination is made as to whether a specified operation is in progress in response to determining that the processor utilization at the first processor exceeds the utilization threshold. The first processor sends a message to the second processor to cause the second processor to reduce a rate at which requests are transferred from the second processor to the first processor in response to determining that the specified operations is in progress.

Abstract: A method for concurrently erasing data on a processor and preparing the processor for removal from a computing system is disclosed. In one embodiment, such a method includes determining tasks queued to be executed on a processor and reassigning the tasks to a different processor, such as to a different processor in the same cluster as the processor. The method further prevents new tasks from being assigned to the processor. The method waits for currently executing tasks on the processor to complete. Once the currently executing tasks are complete, the method initiates a cache-hostile job on the processor to evict entries in cache of the processor. Once the cache-hostile job is complete, the method enables the processor to be removed from a computing system such as a storage system controller. A corresponding system and computer program product are also disclosed.

Abstract: Provided are a computer program product, system, and method for managing I/O requests to a storage array of storage devices in a machine having a processor node and device adaptor. In response to initiating a rebuild of data in the storage array, the device adaptor determines whether a remaining fault tolerance at the storage array comprises a non-zero fault tolerance that permits at least one further storage device to fail and still allow recovery of data stored in the storage array. In response to determining that the remaining fault tolerance is a zero fault tolerance that does not permit at least one storage device to fail and allow recovery of data, the device adaptor sends a message to the processor node to cause the processor node to initiate an emergency protocol to terminate a mission critical operation when the processor node is performing the mission critical operation.

Abstract: Provided are a computer program product, system, and method for fixing anomalies in a preserved data structure used to generate a temporary data structure during system initialization. A preserved data structure in persistent storage is used to build a temporary data structure in a memory of the computing system during initialization of the computing system. The temporary data structure represents computational resources in the computing system and is rebuilt from the preserved data structure during the initialization. The preserved data structure and the temporary data structure are processed to determine whether the preserved data structure includes at least one anomaly that would result in rebuilding the temporary data structure with an error. Information on the preserved data structure and the temporary data structure having the anomaly are processed to determine modifications to correct the preserved data structure. The determined modifications are processed to correct the preserved data structure.

Abstract: Provided are a computer program product, system, and method for reducing a rate at which data is mirrored from a primary server to a secondary server. A determination is made as to whether a processor utilization at a processor managing access to the secondary storage exceeds a utilization threshold. If so, a determination is made as to whether a specified operation at the processor is in progress. A message is sent to the primary server to cause the primary server to reduce a rate at which data is mirrored from the primary server to the secondary server in response to determining that the specified operation is in progress.

Abstract: Provided are a computer program product, system, and method for sharing alias addresses among logical devices for a control unit managing access by hosts to logical devices configured with capacity from attached physical devices. An alias management group of logical devices and alias addresses assigned to the logical devices is configured. A plurality of requests to establish an association of the host with a logical device and the alias addresses assigned to the logical devices in the alias management group are received from a host. Acknowledgment is made to the host that the association is established in response to determining that the host is assigned the logical devices and alias addresses of the logical devices in the alias management group. The host can use one available alias address assigned to any one of the logical devices to access any one of the logical devices indicated in the association.

Abstract: A machine learning module receives inputs comprising attributes of a storage controller, where the attributes affect failures that occur in the storage controller. In response to a failure occurring in the storage controller, a plurality of output values corresponding to a plurality of recovery mechanisms to recover from the failure in the storage controller are generated via forward propagation through a plurality of layers of the machine learning module. A margin of error is calculated based on comparing the generated output values to expected output values corresponding to the plurality of recovery mechanisms, where the expected output values are generated from an indication of a correct recovery mechanism for the failure. An adjustment is made of weights of links that interconnect nodes of the plurality of layers via back propagation to reduce the margin of error, to improve a determination of a recovery mechanism for the failure.

Abstract: In response to an occurrence of a failure in a storage controller, an input on a plurality of attributes of the storage controller at a time of occurrence of the failure is provided to a machine learning module. In response to receiving the input, the machine learning module generates a plurality of output values corresponding to a plurality of recovery mechanisms to recover from the failure in the storage controller. A recovery is made from the failure in the storage controller, by applying a recovery mechanism whose output value is greatest among the plurality of output values that are generated by the machine learning module.

Abstract: Embodiments of the present disclosure may relate to methods and a computer program product for allowing writes based on a granularity level. The method for a storage server may include receiving a received granularity level for a particular volume of a storage device of a client computer including an effective duration for the received granularity level. The method may include receiving an anticipated write to the particular volume at an anticipated write granularity level. The method may include verifying whether the anticipated write granularity level substantially matches the received granularity level at the effective duration. The method may also include writing, in response to the anticipated write granularity level substantially matching the received granularity level at the effective duration, the anticipated write to the particular volume for the received granularity level.