Seamless migration of storage volumes between storage arrays

• US 10,678,465 B2
• Filed: 05/23/2018
• Issued: 06/09/2020
• Est. Priority Date: 04/28/2018
• Status: Active Grant

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Inventors: Feng, Changyu, Li, Liam Xiongcheng, Qiu, Harriet Zhihui, Gao, Hongpo, Liu, Yousheng
• Primary Examiner(s): Warren, Tracy A
• Application Number:
  US15/987,459
• Publication Number:
  US 20190332308A1
• Time in Patent Office: 748 Days
• CPC Class Codes: G06F 3/0604 Improving or facilitating administration, e.g. storage management
• CPC Class Codes: G06F 3/0613 in relation to throughput
• CPC Class Codes: G06F 3/0635 by changing the path, e.g. traffic rerouting, path reconfiguration
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

While migrating a source volume to a destination storage array, disabling the source volume causes a source storage array to send a non-standard failure status code to all host computers in a cluster that includes the source storage array and the destination storage array. While the source volume is disabled with regard to standard host I/O requests received from host computers in the cluster, in response to receipt of the non-standard failure status code, a customized plugin added to the host computers transfers persistent reservation information to the destination storage array using non-standard read and set persistent reservation information commands. The destination volume is then enabled to process host I/O requests in accordance with the transferred persistent reservation information, and an active path for accessing the source volume is modified in each one of the host computers in the cluster to indicate the destination volume in the destination storage array.

Claims

1. A method of providing seamless migration of a source volume located in a source storage array to a destination storage array, comprising the steps of:

disabling the source volume, wherein disabling the source volume prevents the source volume from being used to process standard host I/O requests and causes the source storage array to send a non-standard failure status code to a plurality of host computers in a cluster that also includes the source storage array and the destination storage array;

transferring, by at least one of the host computers in the cluster in response to receipt of the non-standard failure status code, and while the source volume is disabled and prior to enabling a destination volume in the destination storage array to process host I/O requests, persistent reservation information to the destination storage array, wherein the persistent reservation information indicates which ones of the host computers in the cluster are permitted to access the source volume;

in response to determining that the persistent reservation information has been successfully transferred to the destination storage array, enabling the destination volume to process host I/O requests in accordance with the transferred persistent reservation information; and

in response to determining that the destination volume has been enabled to process host I/O requests, modifying an active path for accessing the source volume in each one of the host computers in the cluster to indicate the destination volume, wherein the modification of the active path for accessing the source volume causes subsequent I/O requests that are directed to the source volume to be sent from the host computers in the cluster to the destination storage array for processing in the destination storage array using the destination volume.

2. The method of claim 1, further comprising:

prior to disabling the source volume, sending a hold I/O message to all host computers in the cluster, wherein the hold I/O message causes the host computers in the cluster to stop issuing host I/O requests that are directed to the source volume to the source storage array and to instead create an internal queue in each one of the host computers in the cluster storing host I/O requests that are directed to the source volume; and

in response to determining that the active path for accessing the source volume in each one of the host computers in the cluster has been modified to indicate the destination volume, sending a resume I/O message to all the host computers in the cluster, wherein the resume I/O message causes the host computers in the cluster to resume issuing host I/O requests that are directed to the source volume using the modified active path for accessing the source volume, including the host I/O requests directed to the source volume that are stored in the internal queue in each one of the host computers in the cluster.

3. The method of claim 2, wherein the host computers in the cluster each include a multi-pathing component and a customized plugin component added to the multi-pathing component;

and

wherein transferring the persistent reservation information to the destination storage array by at least one of the host computers in the cluster comprises the customized plugin component added to the multi-pathing component in at least one of the host computers in the cluster transferring the persistent reservation information from the source storage array to the destination storage array.

4. The method of claim 3, wherein the persistent reservation information further indicates one of the host computers in the cluster that is currently permitted to write to the source volume.

5. The method of claim 4, wherein transferring the persistent reservation information from the source storage array to the destination storage array further comprises the customized plugin component added to the multi-pathing component in at least one of the host computers in the cluster obtaining the persistent reservation information from the source storage array at least in part by sending a non-standard read persistent reservations command to the source storage array.

6. The method of claim 5, wherein transferring the persistent reservation information from the source storage array to the destination storage array further comprises the customized plugin component added to the multi-pathing component in at least one of the host computers in the cluster sending the persistent reservation information obtained from the source storage array to the destination storage array by sending a non-standard set persistent reservations command to the source storage array.

7. The method of claim 6, wherein modifying the active path to the source volume in each one of the host computers in the cluster to indicate the destination volume comprises modifying an active path associated with the source volume in the customized multi-pathing component added to the multi-path component in each one of the host computers in the cluster.

8. The method of claim 7, wherein the destination storage array further includes a migration controller component;

and

wherein sending the hold I/O message to all host computers in the cluster comprises the migration controller component in the destination storage array sending the hold I/O messages to all host computers in the cluster.

9. The method of claim 8, wherein disabling the source volume comprises the migration controller component in the destination storage array sending a message to the source storage array that causes the source storage array to prevent the source volume from processing standard host I/O requests from the host computers in the cluster.

10. The method of claim 9, wherein disabling the source volume further causes the source storage array to abort all outstanding host I/O requests in the source storage array that are directed to the source volume and send an I/O failed message to all the host computers in the cluster.

11. The method of claim 10, further comprising:

creating a proxy volume for the destination volume in the destination storage array, wherein the proxy volume for the destination volume processes host I/O requests sent from the host computers in the cluster to the destination storage array and directed to the destination volume at least in part by performing write I/O requests in the host I/O requests sent from the host computers in the cluster to the destination storage array on both the destination volume and the source volume.

12. The method of claim 11, further comprising:

after sending the resume I/O message to all the host computers in the cluster, beginning to copy data stored in the source volume to the destination volume; and

when all data has been copied from the source volume to the destination volume, removing the proxy volume from the destination storage array so that host I/O requests sent from the host computers in the cluster to the destination storage array and directed to the destination volume are subsequently processed by the destination storage array directly on the destination volume.

13. A non-transitory, computer-readable medium including instructions which, when executed by processing units, cause the processing units to perform a method of providing seamless migration of a source volume located in a source storage array to a destination storage array, the method comprising the steps of:

disabling the source volume, wherein disabling the source volume prevents the source volume from being used to process standard host I/O requests and causes the source storage array to send a non-standard failure status code to a plurality of host computers in a cluster that also includes the source storage array and the destination storage array;

transferring, by at least one of the host computers in the cluster in response to receipt of the non-standard failure status code, and while the source volume is disabled and prior to enabling a destination volume in the destination storage array to process host I/O requests, persistent reservation information to the destination storage array, wherein the persistent reservation information indicates which ones of the host computers in the cluster are permitted to access the source volume;

in response to determining that the persistent reservation information has been successfully transferred to the destination storage array, enabling the destination volume to process host I/O requests in accordance with the transferred persistent reservation information; and

in response to determining that the destination volume has been enabled to process host I/O requests, modifying an active path for accessing the source volume in each one of the host computers in the cluster to indicate the destination volume, wherein the modification of the active path for accessing the source volume causes subsequent I/O requests that are directed to the source volume to be sent from the host computers in the cluster to the destination storage array for processing in the destination storage array using the destination volume.

14. The non-transitory, computer-readable medium of claim 13, wherein the method of providing seamless migration of a source volume located in a source storage array to a destination storage array that is caused to be performed by the processing units when the instructions are executed by the processing units further comprises:

prior to disabling the source volume, sending a hold I/O message to all host computers in the cluster, wherein the hold I/O message causes the host computers in the cluster to stop issuing host I/O requests that are directed to the source volume to the source storage array and to instead create an internal queue in each one of the host computers in the cluster storing host I/O requests that are directed to the source volume; and

in response to determining that the active path for accessing the source volume in each one of the host computers in the cluster has been modified to indicate the destination volume, sending a resume I/O message to all the host computers in the cluster, wherein the resume I/O message causes the host computers in the cluster to resume issuing host I/O requests that are directed to the source volume using the modified active path for accessing the source volume, including the host I/O requests directed to the source volume that are stored in the internal queue in each one of the host computers in the cluster.

15. The non-transitory, computer-readable medium of claim 14, wherein the method of providing seamless migration of a source volume located in a source storage array to a destination storage array that is caused to be performed by the processing units when the instructions are executed by the processing units further comprises:

wherein transferring the persistent reservation information to the destination storage array by at least one of the host computers in the cluster comprises the customized plugin component added to the multi-pathing component in at least one of the host computers in the cluster transferring the persistent reservation information from the source storage array to the destination storage array.

16. The non-transitory, computer-readable medium of claim 15, wherein the persistent reservation information further indicates one of the host computers in the cluster that is currently permitted to write to the source volume.

17. The non-transitory, computer-readable medium of claim 16, wherein the method of providing seamless migration of a source volume located in a source storage array to a destination storage array that is caused to be performed by the processing units when the instructions are executed by the processing units further comprises:

wherein transferring the persistent reservation information from the source storage array to the destination storage array further comprises the customized plugin component added to the multi-pathing component in at least one of the host computers in the cluster obtaining the persistent reservation information from the source storage array at least in part by sending a non-standard read persistent reservations command to the source storage array.

18. The non-transitory, computer-readable medium of claim 17, wherein the method of providing seamless migration of a source volume located in a source storage array to a destination storage array that is caused to be performed by the processing units when the instructions are executed by the processing units further comprises:

wherein transferring the persistent reservation information from the source storage array to the destination storage array further comprises the customized plugin component added to the multi-pathing component in at least one of the host computers in the cluster sending the persistent reservation information obtained from the source storage array to the destination storage array by sending a non-standard set persistent reservations command to the source storage array.

19. The non-transitory, computer-readable medium of claim 18, wherein the method of providing seamless migration of a source volume located in a source storage array to a destination storage array that is caused to be performed by the processing units when the instructions are executed by the processing units further comprises:

creating a proxy volume for the destination volume in the destination storage array, wherein the proxy volume for the destination volume processes host I/O requests sent from the host computers in the cluster to the destination storage array and directed to the destination volume at least in part by performing write I/O requests in the host I/O requests sent from the host computers in the cluster to the destination storage array on both the destination volume and the source volume;

after sending the resume I/O message to all the host computers in the cluster, beginning to copy data stored in the source volume to the destination volume; and

when all data has been copied from the source volume to the destination volume, removing the proxy volume from the destination storage array so that host I/O requests sent from the host computers in the cluster to the destination storage array and directed to the destination volume are subsequently processed by the destination storage array directly on the destination volume.

20. A system for providing seamless migration of a source volume, comprising:

a source storage array containing the source volume, the source storage array including memory storing instructions for execution on one or more processing units, and wherein the source storage array is configured and arranged to disable the source volume to prevent the source volume from being used to process standard host I/O requests, and to send a non-standard failure status code to a plurality of host computers that are located in a cluster that also includes the source storage array and a destination storage array;

at least one of the host computers in the cluster configured and arranged to, in response to receipt of the non-standard failure status code, and while the source volume is disabled and prior to enabling a destination volume in the destination storage array to process host I/O requests, transfer persistent reservation information to the destination storage array, wherein the persistent reservation information indicates which ones of the host computers in the cluster are permitted to access the source volume;

wherein the destination storage array is configured and arranged to, in response to a determination that the persistent reservation information has been successfully transferred to the destination storage array, enable the destination volume to process host I/O requests in accordance with the transferred persistent reservation information; and

wherein each one of the host computers in the cluster is configured and arranged to, in response to a determination that the destination volume has been enabled to process host I/O requests, modify an active path for accessing the source volume to indicate the destination volume, wherein the modification of the active path for accessing the source volume causes the host computers in the cluster to send subsequent I/O requests that are directed to the source volume from the host computers in the cluster to the destination storage array for processing in the destination storage array using the destination volume.