Abstract: A data storage system has two computers. Each computer is assigned to a set of data. Two copies of each set of data are maintained. A first copy is stored on a first set of disks and a second copy is stored on a second set of disks. Each time that a data is written by a computer, a label is written to each set of disks, the label having fields for a status of each computer, a first ordinal which is increased each time that a new data is written, and a time stamp giving a time at which the last write was performed. After failure of a computer, a processor determines, in response to reading the labels of the first set of disks and the second set of disks, the most up to date copy of the data assigned to the failed computer.

Abstract: A system and method for storing and retrieving file system log information in a cluster computer system is provided. In response to an aborted takeover operation, a surviving storage system stores a set of file replay log information on storage devices associated with the surviving storage system. When the failed storage system re-initializes, it retrieves the file replay log information from the surviving storage system and utilizes the file replay log information to render its disks consistent.

Abstract: A data storage system has two computers. Each computer is assigned to a set of data. Two copies of each set of data are maintained. A first copy is stored on a first set of disks and a second copy is stored on a second set of disks. Each time that a data is written by a computer, a label is written to each set of disks, the label having fields for a status of each computer, a first ordinal which is increased each time that a new data is written, and a time stamp giving a time at which the last write was performed. After failure of a computer, a processor determines, in response to reading the labels of the first set of disks and the second set of disks, the most up to date copy of the data assigned to the failed computer.

Abstract: A system and method for managing a time-limited long-running process that acts upon disks of a disk array is based upon a general rule of acting upon each disk in an order based upon the length of time from when it was last acted-upon. Disks having the greatest last acted-upon time value (e.g. were processed the longest-time ago) are queued first by the process for acting-upon in the present run. The greatest last acted-upon time includes disks that have no prior last acted-upon time (e.g. newly added/mounted disks). A registry key entry is maintained in a registry file for each of the RAID groups that includes the volume's file system identifier (FSID) and particular RAID group identifier (RGID). The RGID includes the last acted-upon time for the RAID group. This is set to a positive time number if the process has last acted-upon the group to completion.

Abstract: A method for managing a long-running process carried out upon a plurality of disks is disclosed. A registry is established, the registry having a plurality of entries, each entry corresponding to one of the plurality of disks, each entry having a value indicative of a respective time at which its corresponding disk was last acted-upon by the long-running process. The long-running process executes on each of the disks based upon an order in which the disk having an oldest last acted-upon time is processed first and the disk having the newest last acted-upon time is processed last.

Abstract: A technique includes a set of procedures utilized during the transfer of service of a failed storage system to a surviving storage system of a cluster failover (CFO) configuration during a system outage, such as a disaster scenario. The procedures are executed by a RAID subsystem of the surviving storage system during disaster takeover, storage healing and node restart operations. The procedures prevent the possibility that clients accessing the storage system will see out-of-date versions of the data either during transfer of service of the system during the disaster scenario or subsequently after the storage system is healed and service on the failed system has been restored.

Abstract: A system and method for managing a time-limited long-running process that acts upon disks of a disk array is based upon a general rule of acting upon each disk in an order based upon the length of time from when it was last acted-upon. Disks having the greatest last acted-upon time value (e.g. were processed the longest-time ago) are queued first by the process for acting-upon in the present run. The greatest last acted-upon time includes disks that have no prior last acted-upon time (e.g. newly added/mounted disks). A registry key entry is maintained in a registry file for each of the RAID groups that includes the volume's file system identifier (FSID) and particular RAID group identifier to (RGID). The RGID includes the last acted-upon time for the RAID group. This is set to a positive time number if the process has last acted-upon the group to completion.