摘要:
A system of controlling tape drives within a tape drive library where a backup server utilizes client backup schedules and pending client restore requests to efficiently control the powering on and off of tape drives within a tape drive library.
摘要:
A grid application and storage system to enhance data access. The grid application and storage system comprises a plurality of grid application storage nodes with at least one storage system, said grid application storage nodes being remote to each other and connected via inter-connections; at least one application process distributed across said grid application storage nodes; and at least one grid node control process also distributed across said grid application storage nodes, said grid node control process managing grid application storage node operations. A plurality of client systems are connected to said grid application and storage system via a network for deploying said application process and accessing application data stored on said grid application storage nodes.
摘要:
A method and infrastructure are provided for monitoring and analyzing the storage media quality and the drive hardware performance of a media library. The media library comprises a plurality of removable storage media, a plurality of drive units for said storage media and means for moving said storage media to and from appropriate drive units. The media library is managed by a management system being connected to said media library. At least one reference medium of a known and proven quality to compare the performance of at least one of said storage media in a given drive unit with the performance of said reference medium in said given drive unit and/or to test the performance of at least one given drive unit of said media library when acting on said reference medium.
摘要:
A method of controlling tape drives within a tape drive library where a backup server utilizes client backup schedules and pending client restore requests to efficiently control the powering on and off of tape drives within a tape drive library.
摘要:
A load balancing method and system for identifying an input/output (I/O) network path from a set off I/O network paths is provided by the invention. The set off I/O network paths connect a host system via a network to a storage subsystem. The host system comprises at least one host bus adapter (HBA) and the storage subsystem comprises at least one I/O device and the network comprises at least one network device. Each of the HBA, the I/O device and the network device comprise at least one I/O port. For each I/O port of each HBA, an HBA port limit is determined. Additionally the set of I/O network paths which connect the I/O port of each of the HBA via the I/O ports of the network device to the I/O port of the I/O device is identified. Then a fabric utilization limit is determined for each I/O network path and a HBA port utilization is determined for each I/O port of the at least one HBA. All network paths are discarded for which the HBA port utilization is greater than the HBA port limit. For each of the remaining paths a network path distance is determined. All I/O network paths for which the network path distance is greater than the path distance limit are discarded. Then for each of the remaining paths a fabric utilization is determined. All I/O network paths for which the fabric utilization is greater than the fabric utilization limit are discarded and the I/O network path is determined from the remaining network paths.
摘要:
The present invention relates to backup solutions in electronic computing systems and in particular to a method and respective system for managing the storage of application data on a removable storage medium and mounting the removable medium on a according driver device, wherein the application data is cached in a so-called “virtual tape system”, represented by a random-access storage medium, preferably a hard disk, before being written to removable medium or read from removable medium. In order to provide a method including an improved removable medium mount control for increasing the efficiency of removable medium driver device, it is proposed to perform the steps of: managing mount-specific meta data characteristic for removable medium operation workload tasks; predicting upcoming I/O workload based on said meta data; determining based on said calculation, if or when an incoming mount request for mounting a removable medium will be serviced.
摘要:
Various embodiments for optimized power savings in a storage virtualization system are provided. First meta data for physical resources which describes a power status of a storage resource in one of a powered-on/read-write, powered-on/read only, and powered-off power state is created. Second meta data for each of the physical storage resources which determines an actual performance of the physical storage resources and which supports optimization of a powering-on and a powering-off of the physical storage resources is created. A write request from one of a host and application to logical and virtual storage resource is executed.
摘要:
Various embodiments for an Input/Output (I/O) path selection based on workload types are provided. A port workload controller is adapted to control the workload types of ports for accessing a storage device. A storage access command is received by the port workload controller from an application being executed on a virtual server on a physical server. The workload types related to the storage access command are determined. Storage access paths between the application and the storage device having the same workload types as the determined workload types related to the storage access command are determined. Storage access paths mapping data that is controlled by the port workload controller is used. A selected storage access path is selected out of the determined storage access paths having the least utilization. The storage device is accessed via the selected one of the plurality of storage access paths.
摘要:
A method to select a deduplication protocol for a data storage library comprising a plurality of data storage devices configured as a RAID array, by establishing a normal deduplication protocol, a RAID failure deduplication protocol, and a multiple storage device failure deduplication protocol. The method receives host data comprising a plurality of interleaved data blocks. If the system is operating without any storage device failures, then the method processes the host data using the normal deduplication protocol. If the system is operating with a storage device failure, then the method processes the host data using the RAID failure deduplication protocol. If the system is operating with multiple storage device failures, then the method processes the host data using the multiple storage device failure deduplication protocol.
摘要:
A mirrored storage system for applications is provided, which enables and supports the variation and dynamic adaptation of the Recovery Point Objectives (RPO) based on policies. Furthermore, methods are provided for running such a mirrored storage system. Said mirrored storage system comprises a first storage system and at least one further storage system, wherein said first and said further storage system are connected via at least one mirror link. An application accesses said mirrored storage system via a network. Therewith, the data to be stored as response to a write command of said application can be mirrored according to a configurable time-varying RPO requirement of the application transmitting the corresponding write command.