Verified write command in active-active replication
A method is provided, comprising: receiving, at a first storage array, a first request to perform an atomic write operation at a specified address in a volume; placing a read-write lock on a first instance of the volume, and executing the atomic write operation on the first instance of the volume; transmitting, to a second storage array, a second request to perform the atomic write operation on a second instance of the volume; and detecting whether an error condition is present that is associated with the second request; when the error condition is present, comparing a first data item that is stored at the specified address in the first instance of the volume with a second data item that is stored at the specified address in the second instance of the volume, and, when the first data item matches the second data item, transmitting an acknowledgment.
1. A method for use in an active-active storage cluster having first a storage array and a second storage array, the method comprising:
receiving, at the first storage array, a first request to perform an atomic write operation at a specified address in a volume, the first request being received from a host device;
placing a read-write lock on a first instance of the volume that is stored at the first storage array, and executing the atomic write operation on the first instance of the volume;
transmitting, to the second storage array, a second request to perform the atomic write operation on a second instance of the volume that is stored at the second storage array;
detecting whether an error condition is present that is associated with the second request; and
when the error condition is present, comparing a first data item that is stored at the specified address in the first instance of the volume with a second data item that is stored at the specified address in the second instance of the volume, and, when the first data item matches the second data item, transmitting an acknowledgment to the host device indicating that the first request has been completed successfully.
2. The method of claim 1 , further comprising, when the error condition is not present, transmitting an acknowledgment to the host device indicating that the first request has been completed successfully.
3. The method of claim 1 , when the first data item does not match the second data item, storing the second data item, at the specified address, in the first instance of the volume, and transmitting an error message indicating that the first request could not be completed successfully.
4. The method of claim 1 , further comprising releasing the read-write lock after an acknowledgment or an error message is transmitted to the host device.
5. The method of claim 1 , wherein the atomic write operation includes a test and set command.
6. The method of claim 1 , wherein the atomic write operation includes a read-modify-write request.
7. The method of claim 1 , wherein the atomic write request includes a request to change a state of an object that is shared between the host device and at least one other host device that is connected to the active-active storage cluster.
8. A storage system, comprising:
one or more storage devices, and
at least one processor configured to perform the operations of:
receiving a first request to perform an atomic write operation at a specified address in a volume, the first request being received from a host device;
placing a read-write lock on a first instance of the volume that is stored in the one or more storage devices, and executing the atomic write operation on the first instance of the volume;
transmitting, to a peer storage system, a second request to perform the atomic write operation on a second instance of the volume that is stored at the peer storage system;
detecting whether an error condition is present that is associated with the second request; and
when the error condition is present, comparing a first data item that is stored at the specified address in the first instance of the volume with a second data item that is stored at the specified address in the second instance of the volume, and, when the first data item matches the second data item, transmitting an acknowledgment to the host device indicating that the first request has been completed successfully.
9. The storage system of claim 8 , wherein the at least one processor is further configured to perform the operation of, when the error condition is not present, transmitting an acknowledgment to the host device indicating that the first request has been completed successfully.
10. The storage system of claim 8 , wherein the at least one processor is further configured to perform the operation of, when the first data item does not match the second data item, storing the second data item, at the specified address, in the first instance of the volume, and transmitting an error message indicating that the first request could not be completed successfully.
11. The storage system of claim 8 , wherein the at least one processor is further configured to perform the operation of releasing the read-write lock after an acknowledgment or an error message is transmitted to the host device.
12. The storage system of claim 8 , wherein the atomic write operation includes a test and set command.
13. The storage system of claim 8 , wherein the atomic write operation includes a read-modify-write request.
14. The storage system of claim 8 , wherein storage system and the peer storage system are part of an active-active storage cluster, and the atomic write request includes a request to change a state of an object that is shared between the host device and at least one other host device that is connected to the active-active storage cluster.
15. A method for use in an active-active storage cluster having first a storage array and a second storage array, the method comprising:
receiving, at the first storage array, a first request to perform an atomic write operation at a specified address in a volume, the first request being received from a host device;
placing a read-write lock on a first instance of the volume that is stored at the first storage array;
transmitting, to the second storage array, a second request to perform the atomic write operation on a second instance of the volume that is stored at the second storage array; and
detecting whether an error condition is present that is associated with the second request;
when the error condition is present, comparing a first data item that is stored at the specified address in the first instance of the volume with a second data item that is stored at the specified address in the second instance of the volume, and, when the first data item does not match the second data item, copying the second data item to the specified address in the first instance of the volume, and transmitting an acknowledgment to the host device indicating that the first request has been completed successfully.
16. The method of claim 15 , further comprising, when no error condition is present, executing the atomic write operation on the first instance of the volume and transmitting an acknowledgment to the host device indicating that the first request has been completed successfully.
17. The method of claim 15 , when the first data item matches the second data item, transmitting an error message indicating that the first request could not be completed successfully.
18. The method of claim 15 , releasing the read-write lock after an acknowledgment or an error message is transmitted to the host device.
19. The method of claim 15 , wherein the atomic write operation includes a test and set command.
20. The method of claim 15 , wherein the atomic write operation includes a read-modify-write request.