초록 ▼

A method and system for promoting a snapshot in a distributed striped volume system is provided. A master volume server is configured with a rollback process such that when it is determined that a rollback is required, the master volume server sets a flag persistently in its own raid label on disk.

A method and system for promoting a snapshot in a distributed striped volume system is provided. A master volume server is configured with a rollback process such that when it is determined that a rollback is required, the master volume server sets a flag persistently in its own raid label on disk. After the persistent flag is set, the master volume server determines a “common snapshot,” and starts the process of sending RPC messages to each node hosting constituent volumes instructing each constituent volume to roll back to the identified snapshot. When the nodes receive this message a flag is set in the own raid label of each constituent volume and the volume then promotes the particular snapshot. If the master volume server has not received a successful response from each node that the snapshot promotion was successful within a specified time period, there is a retry. The common snapshot is then used as the active file system, thus providing data recovery for the striped volume set.

대표청구항 ▼

What is claimed is: 1. A method of performing data recovery in a striped volume set, comprising: determining a common snapshot for the striped volume set, wherein the common snapshot is a snapshot set that contains a snapshot for each constituent volume in the striped volume set; and in response to

What is claimed is: 1. A method of performing data recovery in a striped volume set, comprising: determining a common snapshot for the striped volume set, wherein the common snapshot is a snapshot set that contains a snapshot for each constituent volume in the striped volume set; and in response to determining the common snapshot, promoting the common snapshot as an active file system for the striped volume set. 2. The method as defined in claim 1, further comprising: detecting that a rollback to the common snapshot is required at a master volume server by receiving a rollback command from an administrator. 3. The method as defined in claim 1, further comprising: periodically polling each constituent volume in the striped volume set; checking to see if a flag is set in a raid label of a constituent volume; in response to the flag in the raid label of the constituent volume being set, setting at a master volume server a flag in its own raid label that a rollback is required; at said master volume server determining the common snapshot to determine a rollback snapshot; and ii notifying each constituent volume that a rollback is required to the common snapshot. 4. The method as defined in claim 3, further comprising: waiting for a response from each constituent volume that the rollback has been successful. 5. The method as defined in claim 4, further comprising: in response to the response not being received from each constituent volume that the rollback has been successful, then said master volume server issuing messages to each constituent volume to retry the rollback to the rollback snapshot. 6. The method as defined in claim 3, further comprising: when the response from each constituent volume that the rollback has been successful has been received by said master volume server, removing at said master volume server, its rollback flag in its own raid label. 7. The method as defined in claim 1 wherein determining a common snapshot, further comprises: sending a query to each constituent volume in the striped volume set to determine snapshot information at that constituent volume; and determining the common snapshot based upon the snapshot information received from one or more constituent volumes. 8. The method as defined in claim 1 wherein determining a common snapshot, further comprises: retrieving snapshot information from each constituent volumes; and determining the common snapshot based upon snapshot information so retrieved. 9. A method of promoting a snapshot in a striped volume set, comprising: periodically obtaining snapshots of each constituent volume in the striped volume set; configuring a master volume server to detect a rollback required state in the striped volume set; in response to detecting the rollback required state, said master volume server setting a flag persistently indicating that a rollback is required; determining a common snapshot for the striped volume set and identifying the common snapshot as a rollback snapshot, wherein the common snapshot is a snapshot set that contains a snapshot for each of the constituent volume in the striped volume set; and in response to determining the common snapshot, notifying constituent volumes in the striped volume set that the rollback snapshot is to be set as an active file system. 10. The method as defined in claim 9, further comprising: receiving responses from each constituent volume that the rollback to the rollback snapshot has been successful; and when the response is received from each constituent volume, removing said flag set persistently by said master volume server. 11. A system for promoting a snapshot in a striped volume set, the system comprising: a master volume server configured to perform a rollback procedure; one or more constituent data volumes configured to store stripes of data of data containers as the striped volume set; a snapshot manager associated with each one of said constituent volumes, said snapshot manager configured to periodically generate snapshots of an active file system; and a rollback process running on said master volume server configured to generate messages for each of said constituent volumes to instruct each node that hosts each volume to promote a common snapshot as the active file system, wherein the common snapshot is a snapshot set that contains a snapshot for each of the constituent volume in the striped volume set. 12. The system of claim 11, further comprising: a raid label in each volume, said raid label including a flag field to identify that a rollback is required. 13. The system of claim 11 wherein each constituent data volume of the striped volume set is served by one or more disk elements. 14. The system of claim 11 wherein the striped volume set is defined by a set striping rules. 15. The system of claim 11 wherein a client interfaces with the system using a network element. 16. A computer readable medium containing executable program instructions executed by a processor, comprising: program instructions that obtain, periodically, snapshots of each constituent volume in a striped volume set; program instructions that configuring a master volume server to detect a rollback required state in the striped volume set; program instructions that set a flag persistently, by said master volume, indicating that a rollback is required, in response to program instructions that detect a rollback required state; program instructions that determine a common snapshot for the striped volume set and identify a common snapshot as a rollback snapshot, wherein the common snapshot is a snapshot set that contains a snapshot for each of the constituent volume in the striped volume set; and program instructions that notify constituent volumes in the striped volume set that the rollback snapshot is to be set as an active file system. 17. The computer readable medium as defined in claim 16, further comprising: program instructions that receive responses from each constituent volume that the rollback to the rollback snapshot has been successful; and program instructions that remove said flag set persistently by said master volume server when a response is received from each constituent volume. 18. The computer readable medium as defined in claim 16, wherein the program instructions that detect the rollback required state at said master volume server receive a rollback command from an administrator. 19. The computer readable medium as defined in claim 16, further comprising: program instructions that poll, periodically, each constituent volume in the striped volume set; program instructions that check to see if a flag is set in a raid label of a constituent volume; program instructions that set, at said master volume server, a flag in its own raid label that a rollback is required in response to the flag being set at the raid label of the constituent volume; program instructions that determine, at said master volume server, the common snapshot to determine a rollback snapshot; and program instructions that notify each constituent volume that the rollback is required to the rollback snapshot. 20. The computer readable medium as defined in claim 16 further comprising: program instructions that wait for a response from each constituent volume that the rollback has been successful; and program instructions that issue messages by said master volume server to each constituent volume to retry the rollback to the rollback snapshot, in response to not receiving from each constituent volume a message that the rollback has been successful. 21. A method of performing data recovery in a striped volume set, comprising: periodically polling each constituent volume in the striped volume set; checking to see if a flag is set in a raid label of a constituent volume; in response to the flag being set in the raid label of the constituent volume, setting at a master volume server a flag in its own raid label that a rollback is required; in response to the rollback being required, determining a common snapshot for the striped volume set, wherein the common snapshot is a snapshot set that contains a snapshot for each constituent volume in the striped volume set; notifying each constituent volume that the rollback is required to the common snapshot; and promoting the common snapshot as an active file system at each constituent volume of the striped volume set. 22. The method as defined in claim 21, further comprising: waiting for a response from each constituent volume that the rollback has been successful. 23. The method as defined in claim 22, further comprising: in response to the response not being received from each constituent volume that the rollback has been successful, then said master volume server issuing messages to each constituent volume to retry the rollback to the rollback snapshot. 24. The method as defined in claim 21, further comprising: when the response from each constituent volume that the rollback has been successful has been received by said master volume server, removing at said master volume server, its rollback flag in its own raid label. 25. The method as defined in claim 21 wherein the determining a common snapshot, further comprises: sending a query to each constituent volume in the striped volume set to determine snapshot information at that constituent volume; and determining the common snapshot based upon the snapshot information received from one or more constituent volumes. 26. The method as defined in claim 21 wherein determining a common snapshot, further comprises: retrieving snapshot information from each constituent volumes; and determining the common snapshot based upon the retrieved snapshot information. 27. A system for promoting a snapshot in a striped volume set, the system comprising: one or more constituent data volumes configured to store stripes of data of data containers as the striped volume set; a snapshot manager associated with each one of said constituent volumes, the snapshot manager configured to periodically generate snapshots of that constituent volume; a master volume server configured to periodically poll each constituent volume in the striped volume set to see if a flag is set in a raid label of each constituent volume, the master volume further configured to set a flag in its own raid label that a rollback is required in response to a flag being set at the raid label of the constituent volume; and a rollback process running on said master volume server configured to generate messages for each constituent volume instructing each node hosting each constituent volume that the rollback is required and to promote a common snapshot as an active file system, wherein the common snapshot is a snapshot set that contains a snapshot for each of the constituent volume in the striped volume set. 28. The system of claim 27 wherein each constituent data volume of the striped volume set is being served by one or more disk elements. 29. The system of claim 27 wherein the striped volume set is defined by a set striping rules. 30. The system of claim 27 wherein a client interfaces with the system using a network element.