Leveraging snapshots

1. A computer implemented method for continuous replication in a virtual environment comprising:

recovering by a computer system, replication of a first volume of a virtual machine running on a hypervisor on a first site to a second site after a crash of the hypervisor of the first site wherein, prior to the crash, the first volume is continuously being replicated to the second site;

determining by the computer system, using an API, differences between the first volume on the first site in its current state and a replication of the first volume from a previous point in time on the second site that corresponds to a point in time prior to the crash; wherein the determined differences represent changes to the first volume of the virtual machine running on the hypervisor of the first site that were lost at the time the hypervisor crashed and were not replicated to the second site; and

using by the computer system the determined differences, restoring replication of the first volume to the second site; wherein restoring replication includes bringing a replication of the first volume on the second site to be consistent with the current state of the first volume of the first site by applying the determined differences to the second site and setting further changes of the first volume to be sent to the second site.

2. The method of claim 1 wherein the previous point in time is a snapshot and the snapshot was created using an API.

3. The method of claim 2 further comprising:

synchronizing an image of the first volume and an image of the second site based on the determined differences.

4. The method of claim 2 wherein the snapshot is created using a VASA API.

5. The method of claim 2 further comprising: waiting for the snapshot to arrive at the second site.

6. The method of claim 5 further comprising:

creating a new snapshot, erasing a previous snapshot; and

renaming the new snapshot to be a previous snapshot.

7. The method of claim 1 further comprising:

adding the determined differences to a delta marking stream.

8. The method of claim 7 further comprising:

creating a new snapshot, erasing a previous snapshot; and

renaming the new snapshot to be a previous snapshot.

9. The method of claim 1 wherein the hypervisor of first site has a data replication appliance and wherein the data replication appliance stores changes to the first volume to be sent to the second site; wherein the crash of the hypervisor causes the data replication appliance to crash and lose stored changes to the first volume to be sent to the second site.

10. The method of claim 1 wherein the continuous replication of the first volume comprises a tracking process to track data blocks that are changed on the first volume but not replicated to the second site and wherein bringing a replication of the first volume on the second site to be consistent with the current state of the first volume comprises synchronizing the changed data blocks that were tracked as not being replicated to the second site.

11. A computer program product for use in replication comprising:

a non-transitory computer readable medium encoded with computer executable program code for replication of data, which when executed by one or more processors cause the one or more processors to perform a method comprising the steps of:

recovering, replication of a first volume of a virtual machine running on a hypervisor on a first site to a second site after a crash of the hypervisor of the first site wherein, prior to the crash, the first volume is continuously being replicated to the second site;

determining, using an API, the differences between the first volume on the first site in its current state and a replication of the first volume from a previous point in time on a second site that corresponds to a point in time prior to the crash; wherein the first volume is being replicated to the second site; wherein the determined differences represent changes to the first volume of the virtual machine running on the hypervisor of the first site that were lost at the time the hypervisor crashed and were not replicated to the second site; and

using the determined differences, restoring replication of the first volume to second site; wherein restoring replication includes brining a replication of the first volume on the second site to be consistent with the current state of the first volume of the first site by applying the determined differences to the second site and setting further changes of the first volume to be sent to the second site.

12. The computer program product of claim 11 wherein the previous point in time is a snapshot and the snapshot was created using an API.

13. The computer program product of claim 12 the code further configured to cause the execution of:

waiting for the snapshot to arrive at the second site.

14. The computer program product of claim 13 , the code further configured to cause the execution of:

creating a new snapshot, erasing the previous snapshot; and

renaming the new snapshot to be the previous snapshot.

15. The computer program product of claim 11 the code further configured to cause the execution of:

adding the determined differences to a delta marking stream.

16. The computer program product of claim 15 the code further configured to cause the execution of:

synchronizing an image of the first volume and an image of the second site based on the determined differences.

17. The computer program product of claim 16 the code further configured to cause the execution of:

creating a new snapshot, erasing the previous snapshot; and

renaming the new snapshot to be the previous snapshot.

18. A system for continuous replication in a virtual environment, the system comprising:

a first site;

an API;

a second site; and

a memory and one or more processors, and computer-executable logic stored in the memory, wherein the computer-executable program logic is programmed to be executed by the one or more processors to perform:

recovering, replication of a first volume of a virtual machine running on a hypervisor on the first site to the second site after a crash of the hypervisor of the first site wherein, prior to the crash, the first volume is continuously being replicated to the second site using a splitter;

determining, using the API, the differences between the first volume on the first site in its current state and a replication of the first volume from a previous point in time on the second site; wherein the determined differences represent changes to the first volume of the virtual machine running on the hypervisor of the first site that were lost at the time the hypervisor crashed and not replicated to the second site; and

using the determined differences, restoring replication of the first volume to second site; wherein restoring replication includes bringing a replication of the first volume on the second site to be consistent with the current state of the first volume of the first site by applying the determined differences to the second site and setting further changes of the first volume to be sent to the second site.

19. The system of claim 18 wherein the previous point in time is a snapshot and the snapshot was created using an API.

20. The system of claim 18 , the program logic further programmed to be executed by the one or more processors to perform-adding the determined differences to a delta marking stream.

21. The system of claim 18 , the program logic further programmed to be executed by the one or more processors to perform

synchronizing an image of the first volume and an image of the second site based on the determined differences.