摘要:
A system for providing high speed external backup and restoration of data in a data storage system. Individual data objects are backed up in logical order on backup media such as magnetic tape so that they may be retrieved individually. Minimized descriptions of backed up data are transmitted between the host and backup systems to control high speed transfer of data between data servers and backup devices on a separate data channel. The descriptions are manipulated by the backup system to comply with formats of any number of host systems and data storage systems and to efficiently map the backed up data onto the backup devices. The system also provides for restoration of data to the data storage system by using a host system to pre-allocate space for the restored data, obtaining mapping information for the allocated space from the host system, and then directly writing the restored data into the allocated space, based on the mapping information.
摘要:
A method of attribute based resource allocation is presented. A resource such as a pool of storage has certain properties or attributes associated with portions of the pool. When a request for storage is needed, the requester provides a selection expression which comprises properties and logical operators. From the selection expression a determination is made to allocate portions of the storage pool which meet the requirements of the selection expression.
摘要:
A method of attribute based resource allocation is presented. A resource such as a pool of storage has certain properties or attributes associated with portions of the pool. When a request for storage is needed, the requester provides a selection expression which includes properties and logical operators. From the selection expression a determination is made to allocate portions of the storage pool which meet the requirements of the selection expression.
摘要:
A method is used in mapping data storage and virtual machines. A logical volume from a data storage system is provided for use by a hypervisor. The hypervisor is queried through a web service to identify a virtual machine of the hypervisor. It is determined that the virtual machine is using the logical volume.
摘要:
Alert generation is managed. A frequency with which a performance metric satisfies a first condition is determined. It is determined whether the frequency satisfies a second condition. An alert is generated based on the determinations.
摘要:
A system for efficiently representing or “mapping” data so that it can be rapidly communicated to a back-up storage system from a primary processor or a shared storage device while at the same time allowing a backup system to backup files rather than devices is described.
摘要:
Mechanisms and techniques are disclosed that provide the ability to backup and restore a file between a backup server and a data storage system using multiple trails of data for the same file. Within the data storage system, the file is generally stored as a series of data portions spread across multiple storage devices such as disk. The file may be stored, for example, within a volume. In the backup operation of the invention, a discover phase determines which respective data portions of the file (e.g., partitions) are stored upon which respective storage devices (e.g., disks) of a plurality of storage devices. The system then designates individual respective backup devices (e.g., tapes) of a plurality of backup devices to be used for backing-up respective individual ones of the data portions of the file that are stored on respective storage devices. Next, the system concurrently backs up at least two respective data portions of the file to different respective backup devices for which those data portions are designated, such that at least two respective data portions of the file are backed-up to respective backup devices in parallel. The restore operation operates in a similar manner by determining which respective data portions (e.g., backup images) of the file are stored upon which respective backup devices (e.g., tapes) of a plurality of backup devices. The system then designates individual respective portions of storage devices of a plurality of storage devices to be used for storing respective individual ones of the data portions of the file that are stored on respective backup devices. Next, the system concurrently restores at least two respective data portions of the file to different respective storage devices for which those data portions are designated, such that at least two respective data portions of the file are restored to respective storage devices from respective backup devices in parallel. By treating portions of a file as separate backup work items, each portion may be provided as an independent stream or trail of data to or from the backup server, thus decreasing the time required to backup or restore a file.
摘要:
A hierarchical storage management system for database management systems that divides a database logically into separately managed regions, with each region being described by an entry in a vector kept in a regions file. The region entry contains a time stamp of the last time the region was accessed, the staging identifier of the region if it has been migrated, the base level backup staging identifier of the region if it has been baselined, and an indicator telling whether or not the region is resident online. Each managed region of the database is migrated to a separate staging file. When the database software issues a read or write input/output operation, the present invention sends the migration software a signal signifying this has occurred. The migration software of the present invention then updates the accessed time stamp, and checks to see if the region is resident. If it is not resident, it is staged in. The present invention creates and updates a migration candidate list ordered by date last accessed and region size. Both demand staging by the HSM, and user--initiated staging can then operate on the migration candidate list to migrate suitable files to tertiary storage.
摘要:
Described are techniques for event notification. A first notification about a first event is received. A timer is set to an amount of time. Processing waits until an occurrence of either receiving a second notification of a second event or expiration of said timer indicating that the amount of time has lapsed. It is determined whether the amount of time has lapsed or whether the second notification has been received. If the expiration time has lapsed, notification of the first event is provided, and otherwise, if the second notification has been received prior to the amount of time lapsing, the timer is reset and processing waits for an occurrence of either receiving a next notification of an event or expiration of the timer indicating that the amount of time has lapsed.
摘要:
A technique, which involves participation by electronic circuitry of a data storage array, obtains configuration change information from host devices which are constructed and arranged to store data into and load data from the data storage array. In particular, the technique involves receiving a set of change notices from a set of host devices. Each change notice is received by the electronic circuitry of the data storage array when a respective host device of the set of host devices undergoes a configuration change. The technique further involves providing a set of requests from the electronic circuitry of the data storage array to the set of host devices based on the set of change notices. The technique further involves acquiring configuration change information from the set of host devices in response to the set of requests.