Consistent replication of virtual computing instance data

1. A non-transitory computer readable storage medium comprising instructions that cause computer system to carry out a method of replicating data for a virtual computing instance wherein the virtual computing instance and a virtualization management module are running on hypervisor executing on a host computer communicatively coupled to a storage system, said method comprising:

determining, by the virtualization management module, a set of logical storage volumes across which the virtual computing instance issues dependent write IOs wherein the logical storage volumes are provisioned in the storage system;

issuing, by the virtualization management module, a first command to the virtual computing instance to block the virtual computing instance from issuing new IOs to the logical storage volumes of the set of logical storage volumes in the storage system, and to block the virtual computing instance from receiving IO acknowledgements sent by the storage system without waiting for in-flight IOs to complete;

responsive to issuing the first command and before issuing a second command to the virtual computing instance wherein the second command unblocks the virtual computing instance from issuing new IOs to the storage system and unblocks the virtual computing instance from receiving IO acknowledgements sent by the storage system, issuing, by the virtualization module, a command to the storage system wherein the command instructs the storage system to create replicas of all the logical storage volumes in the set to which the virtual computing instance issues dependent write IOs on a second storage system such that the dependent writes of the replicas are ordered in the same manner as the logical storage volumes in the set; and

issuing, by the virtualization module, the second command to the virtual computing instance.

2. The non-transitory computer readable storage medium of claim 1 , wherein the logical storage volumes are virtual volumes that include a first virtual volume that stores data of a first virtual disk for the virtual computing instance and a second virtual volume that stores data of a second virtual disk for the virtual computing instance.

3. The non-transitory computer readable storage medium of claim 2 , wherein each of the virtual volumes has an independent storage policy.

4. The non-transitory computer readable storage medium of claim 1 wherein the virtualization module waits for an acknowledgement from the storage system that replicas have been created before issuing the second command.

5. The non-transitory computer readable storage medium of claim 1 , wherein the virtualization management module determines the set of logical storage volumes across which the virtual computing instances issue dependent write IOs in response to the write requests, from consistency group definitions maintained by the virtualization management module.

6. A method of replicating data for a virtual computing instance wherein the virtual computing instance and a virtualization management module are running on hypervisor executing on a host computer communicatively coupled to a storage system, said method comprising:

determining, by the virtualization management module, a set of logical storage volumes across which the virtual computing instances issue dependent write IOs wherein the logical storage volumes are provisioned in the storage system;

issuing, by the virtualization module, a first command to the virtual computing instance to block the virtual computing instance from issuing new IOs to the logical storage volumes of the set of logical storage volumes in the storage system, and to block the virtual computing instance from receiving IO acknowledgements sent by the storage system without waiting for in-flight IOs to complete;

responsive to issuing the first command and before issuing a second command to the virtual computing instance wherein the second command unblocks the virtual computing instance from issuing new IOs to the storage system and unblocks the virtual computing instance from receiving receipt of IO acknowledgements sent by the storage system, issuing, by the virtualization module, a command to the storage system wherein the command instructs the storage system to create replicas of all the logical storage volumes in the set to which the virtual computing instance issues dependent write IOs on a second storage system such that the dependent writes of the replicas are ordered in the same manner as the logical storage volumes in the set; and

issuing the second command to each of the virtual computing instances.

7. The method of claim 6 , wherein the logical storage volumes are virtual volumes that include a first virtual volume that stores data of a virtual disk of one of the virtual computing instances and a second virtual volume that stores data of a virtual disk of another one of the virtual computing instances.

8. The method of claim 7 , wherein each of the virtual volumes has an independent storage policy.

9. The method of claim 6 , wherein virtualization management module determines the set of logical storage volumes across which the virtual computing instances issue dependent write IOs in response to the write requests, from consistency group definitions maintained by the virtualization management module.

10. The method of claim 6 wherein the virtualization management module waits for an acknowledgement from the storage system that replicas have been created before issuing the second command.

11. The method of claim 6 , wherein the creation of replicas comprises using an API to instruct the storage system to flush an IO cache for the logical storage volumes in the set such that write data for the logical storage volumes in the set that has already been acknowledged has been written.

12. The method of claim 6 , wherein the creation of replicas comprises using an API to instruct the storage system to remove one or more replicas in response to determining that an error occurred in creating the replica.

13. A computer system comprising:

a processor; and

a non-transitory computer readable medium having stored thereon program code for replicating data for a virtual computing instance wherein the virtual computing instance and a virtualization management module are running on hypervisor executing on a host computer communicatively coupled to a storage system, the program code causing the processor to:

determine, by the virtualization management module, a set of logical storage volumes across which the virtual computing instance issues dependent write IOs wherein the logical storage volumes are provisioned in the storage system;

issue, by the virtualization module, a first command to the virtual computing instance to block the virtual computing instance from issuing new IOs to the logical storage volumes of the set of logical storage volumes in the storage system, and to block the virtual computing instance from receiving IO acknowledgements sent by the storage system without waiting for in-flight IOs to complete;

responsive to issuing the first command and before issuing a second command to the virtual computing instance wherein the second command unblocks the virtual computing instance from issuing new IOs to the storage system and unblocks the virtual computing instance from receiving IO acknowledgements sent by the storage system, issue, by the virtualization module, a command to the storage system wherein the command instructs the storage system to create replicas of all the logical storage volumes in the set to which the virtual computing instance issues dependent write IOs on a second storage system such that the dependent writes of the replicas are ordered in the same manner as the logical storage volumes in the set; and

issue, by the virtualization module, the second command to the virtual computing instance.

14. The computer system of claim 13 , wherein the logical storage volumes are virtual volumes that include a first virtual volume that stores data of a first virtual disk for the virtual computing instance and a second virtual volume that stores data of a second virtual disk for the virtual computing instance.

15. The computer system of claim 14 , wherein each of the virtual volumes has an independent storage policy.

16. The computer system of claim 13 wherein the virtualization module waits for an acknowledgement from the storage system that replicas have been created before issuing the second command.

17. The system of claim 13 , wherein the virtualization management module determines the set of logical storage volumes across which the virtual computing instances issue dependent write IOs in response to the write requests, from consistency group definitions maintained by the virtualization management module.