Abstract: Consistency errors may arise during replication of diverse data storage objects from servers in a data network. For preventing consistency errors, a server is programmed for responding to a request for a report of whether or not the server can replicate the dataset without causing a consistency error. Such a report is useful for replicating a consistency group of objects from more than one server. Preparation for replication can be begun at each of the servers, and if any of the servers fail to report that replication can be done without causing a consistency error, the preparation can be aborted so that write access by applications will not be disrupted while the potential problem is diagnosed and eliminated.

Abstract: A snapshot copy facility stores snapshot copies of a production, file system. The snapshot copy facility receives a request for the difference between a specified older snapshot,copy and a specified younger snapshot copy, and responds by returning the difference between the older snapshot copy and the younger snapshot copy. In a preferred implementation, the snapshot copy facility has an index for each snapshot copy for indicating blocks of data in the production file system that have changed between the snapshot copy and a next snapshot copy of the production file system. The indices are scanned for a sequence of the snapshot copies to determine the blocks that have changed, and the snapshot copy facility returns the block numbers and data in the younger snapshot copy for the blocks that have changed.

Abstract: Consistent updates are made automatically over a wide-area IP network, concurrently with read-only access to the remote copies. A replication control protocol (RCP) is layered over TCP/IP providing the capability for a remote site to replicate and rebroadcast blocks of the remote copy data to specified groups of destinations, as configured in a routing table. A volume multicast layer over RCP provides for multicasting to specified volume extents of the blocks. The blocks are copied at the logical level, so that it does not matter what physical structure is used for storing the remote copies. Save volumes buffer the remote copy data transmitted between the primary or secondary file system volume and the IP network, in order to ensure independence between the replication process, the IP transport method, and the primary file system being replicated.

Abstract: File systems possibly impacted by a bad sector in a disk drive are unmounted. Then the disk drive is scanned in order to identify additional bad sectors, and for each identified bad sector, a logical address of the bad sector is re-vectored to a good sector. An attempt is made to restore the data of each bad sector, and then the file system is checked for consistency. If a bad sector cannot be restored, any application relying on the bad sector is aborted. Once the file system is found to be consistent, the file system is remounted. Re-vectored blocks are reverse mapped to identify damaged or repaired directories and files. Damaged or repaired directories and files are reported to a system administrator, and damaged directories or files of a client are reported to the client.

Abstract: In an asynchronous remote copy system, a primary site keeps a list of data blocks for which changes have been made to a primary file system and transmitted to a secondary file system at a secondary site. When the primary site becomes inoperative, the secondary site begins read/write access to the secondary file system and keeps a snapshot copy of the restart point. Once the primary site becomes operative, the primary file system is restored to the state of the secondary file system at the restart point by using the list of data blocks for obtaining from the snapshot copy the data of the blocks for which changes had been made to the primary file system after the restart point. Then the primary file system is synchronized to the secondary file system, and read/write access is switched back from the secondary file system to the primary file system.

Abstract: Consistency errors may arise during replication of diverse data storage objects from servers in a data network. For preventing consistency errors, a server is programmed for responding to a request for a report of whether or not the server can replicate the dataset without causing a consistency error. Such a report is useful for replicating a consistency group of objects from more than one server. Preparation for replication can be begun at each of the servers, and if any of the servers fail to report that replication can be done without causing a consistency error, the preparation can be aborted so that write access by applications will not be disrupted while the potential problem is diagnosed and eliminated.

Abstract: A file server maintains a production file system supported by a clone volume, and multiple snapshot file systems supported by respective save volumes in a snapshot queue. Before a data block is modified for the first time after creation of the youngest snapshot, the data block is copied from the clone volume to the save volume of the youngest snapshot. A bit map indicates the blocks that have already been copied, and a block map shows the save block address for each corresponding clone block address. When a new snapshot is created, the bit and block maps are converted to a hash index that is kept linked to the save volume of what had been the youngest snapshot. When other than the oldest snapshot file system is deleted, the respective save volume is retained as a hidden object until it becomes the oldest save volume.

Abstract: In an asynchronous remote copy system, a primary site keeps a list of data blocks for which changes have been made to a primary file system and transmitted to a secondary file system at a secondary site. When the primary site becomes inoperative, the secondary site begins read/write access to the secondary file system and keeps a snapshot copy of the restart point. Once the primary site becomes operative, the primary file system is restored to the state of the secondary file system at the restart point by using the list of data blocks for obtaining from the snapshot copy the data of the blocks for which changes had been made to the primary file system after the restart point. Then the primary file system is synchronized to the secondary file system, and read/write access is switched back from the secondary file system to the primary file system.

Abstract: A snapshot copy facility stores snapshot copies of a production, file system. The snapshot copy facility receives a request for the difference between a specified older snapshot,copy and a specified younger snapshot copy, and responds by returning the difference between the older snapshot copy and the younger snapshot copy. In a preferred implementation, the snapshot copy facility has an index for each snapshot copy for indicating blocks of data in the production file system that have changed between the snapshot copy and a next snapshot copy of the production file system. The indices are scanned for a sequence of the snapshot copies to determine the blocks that have changed, and the snapshot copy facility returns the block numbers and data in the younger snapshot copy for the blocks that have changed.

Abstract: A file server maintains a production file system supported by a clone volume, and multiple snapshot file systems supported by respective save volumes in a snapshot queue. Before a data block is modified for the first time after creation of the youngest snapshot, the data block is copied from the clone volume to the save volume of the youngest snapshot. A bit map indicates the blocks that have already been copied, and a block map shows the save block address for each corresponding clone block address. When a new snapshot is created, the bit and block maps are converted to a hash index that is kept linked to the save volume of what had been the youngest snapshot. When other than the oldest snapshot file system is deleted, the respective save volume is retained as a hidden object until it becomes the oldest save volume.

Abstract: Consistent updates are made automatically over a wide-area IP network, concurrently with read-only access to the remote copies. A replication control protocol (RCP) is layered over TCP/IP providing the capability for a remote site to replicate and rebroadcast blocks of the remote copy data to specified groups of destinations, as configured in a routing table. A volume multicast layer over RCP provides for multicasting to specified volume extents of the blocks. The blocks are copied at the logical level, so that it does not matter what physical structure is used for storing the remote copies. Save volumes buffer the remote copy data transmitted between the primary or secondary file system volume and the IP network, in order to ensure independence between the replication process, the IP transport method, and the primary file system being replicated.