Abstract: An embodiment of the invention provides an apparatus and method for optimizing reads in a mirrored file system. The apparatus and method receives a plurality of read request blocks for reading the target regions in disks. The apparatus and method will chain the read request blocks in a first set, where all of the read request blocks in the first set are targeted to a first disk. The read request blocks in the first set are then sent as a read request to the first disk. When each read request block in the first set has completed accessing their respective target regions in the first disk, then the read data is generated from the first disk. Therefore, by chaining the read request blocks in a read request to a single disk, read striping is not performed across multiple disks during a read operation.

Abstract: An embodiment of the invention provides an apparatus that includes a primary node, a secondary node, and an intermediate node. Data is synchronously copied from the primary node to the intermediate node, and data is asynchronously copied from the intermediate node to the secondary node. The data stored in the intermediate node is a subset of the data set that is stored in the primary node. Therefore, an embodiment of the invention advantageously provides a multiple hop data replication method that allows data recovery if disaster occurs at the location of the primary node. Additionally, an embodiment of the invention advantageously provides a tolerable latency value from the time that a client device sends a write request to the primary node to the time that an acknowledgement message is received by the client device.

Abstract: An embodiment of the invention provides an apparatus that includes a primary node, a secondary node, and an intermediate node. Data is synchronously copied from the primary node to the intermediate node, and data is asynchronously copied from the intermediate node to the secondary node. The data stored in the intermediate node is a subset of the data set that is stored in the primary node. Therefore, an embodiment of the invention advantageously provides a multiple hop data replication method that allows data recovery if disaster occurs at the location of the primary node. Additionally, an embodiment of the invention advantageously provides a tolerable latency value from the time that a client device sends a write request to the primary node to the time that an acknowledgement message is received by the client device.

Abstract: A method for transferring data from a source computer to a destination computer, and restarting the source computer transmission after a halt of the transmission, has the following steps. The source computer inserts a checkpoint number into a data stream, the data flowing in the data stream. The checkpoint number is stored at the source computer. The data stream is halted in response to an error in a path of the data stream. The destination computer locates the last checkpoint number successfully received before halting the data stream. The last checkpoint number successfully received before halting the data stream is transferred to the source computer; and the source computer compares it with checkpoint numbers stored in the source computer to determine where in the data stream to resume transfer of the data. Transfer of the data is resumed in response to the last checkpoint number.

Abstract: An improved method and apparatus for quickly and efficiently updating the original source volume and original target volumes after the original source volume has become temporarily unavailable. The original target volume is characterized as a source volume while the original source volume is temporarily unavailable. Transfer lists of different data blocks are generated. Data blocks not originally found on a source are copied to the target. Data blocks included on a target that were not found on the source are removed. By focusing upon specific data blocks, this technique avoids the use of filer overhead and other computational resources that would be expended if the entire volume were recopied.

Abstract: Propagation of data by mirroring the data from a first filer to a second filer, and then mirroring the data from the second filer to a third filer. More generally, propagating data by mirroring the data along a chain or cascade of mirrors. Snapshots are used in mirroring the data. Preferably, the data is organized in one or more volumes on the filer, and one or more of the volumes are mirrored. Deletion is prevented of a snapshot associated with a volume from a filer until the filer is informed that no other filer needs that snapshot to mirror the volume from the filer. Preferably, each filer uses softlocks to track what snapshots correspond to volumes mirrored down the chain or cascade of filers.

Abstract: An improved method and apparatus for quickly and efficiently updating the original source volume and original target volumes after the original source volume has become temporarily unavailable. The original target volume is characterized as a source volume while the original source volume is temporarily unavailable. Transfer lists of different data blocks are generated. Data blocks not originally found on a source are copied to the target. Data blocks included on a target that were not found on the source are removed. By focusing upon specific data blocks, this technique avoids the use of filer overhead and other computational resources that would be expended if the entire volume were recopied.