Continuous data replication
In a first embodiment, a method and computer program product for use in a storage system comprising quiescing IO commands the sites of an ACTIVE/ACTIVE storage system, the active/active storage system having at least two storage sites communicatively coupled via a virtualization layer, creating a change set, unquiescing IO commands by the virtualization layers, transferring data of a change set to the other sites of the active/active storage system by the virtualization layer, and flushing the data by the virtualization layer. In a second embodiment, a method and computer program product for use in a storage system comprising fracturing a cluster of an active/active storage system; wherein the cluster includes at least two sites, stopping IO on a first site of the cluster; and rolling to a point in time on the first site.
1. 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, the code configured to enable the execution of:
quiescing IO commands to sites of an ACTIVE/ACTIVE storage system, the active/active storage system having at least two storage sites communicatively coupled via a virtualization layer; wherein the virtualization layer enables the at least two storage sites to present a same virtual volume at each of the at least two storage sites enabling transparent migration of a virtual machine between the at least two storage sites; wherein the virtualization layer enables cache coherency between each of the at least two storage sited; wherein the virtualization layer enables simultaneous, active/active, concurrent read/write access to the same virtual volume at both of the sites;
creating a change set for the same virtual volume;
unquiescing IO commands to the sites;
transferring data of the change set to the other sites of the active/active storage system by the virtualization layer, and
flushing the data to the sites; wherein flushing of the data of the change set enables the other sites of the at least two storage sites to synchronize changed data of the site from which the change set was transferred enabling the at least two storage sites to have a point in time with the same data.
2. The code of claim 1 wherein the flushing comprises:
reading, by the virtualization layer, an undo of the data to be flushed in the change set;
writing the undo data corresponding to the data in the write cache to a data protection appliance (DPA);
writing, by the DPA, the undo data to a journal; and
writing the data to the backend storage system by the virtualization layer.
3. The code of claim 2 where the DPA is a virtual machine; and the code further enables:
writing the data from the write cache to a volume;
notifying the DPA a full change set has been flushed; and
creating a bookmark at the undo journal by the DPA.
4. the code of claim 1 for use in replication enabling:
a non-transitory computer readable medium encoded with computer executable program code for replication of data, the code configured to enable the execution of:
fracturing a cluster of an active/active storage system; wherein the cluster includes at least two sites; wherein an active/active storage system that is not fractured includes a virtualization layer that enables the at least two storage sites to present a same virtual volume at each of the at least two storage sites enabling transparent migration of a virtual machine between the at least two storage sites; wherein the virtualization layer enables simultaneous, active/active, concurrent read/write access to the same virtual volume at both of the sites; wherein the virtualization layer enables cache coherency between each of the at least two storage sited; wherein a fractured active/active storage system enables an image on a first site of the sites to be rolled to a point in time and enables active production on a second site of the sites;
stopping IO on a first site of the cluster; and
rolling to a point in time on a first site of the cluster that is consistent with at least a second site of the cluster before the at least two sites were fractured.
5. The code of claim 4 further enabling:
flushing a write cache on a second site of the cluster to the first site of the cluster.
6. The code of claim 4 further enabling:
requesting, by a user, access to a point in time;
writing undo information generated by rolling to the point in time to a redo log;
exposing the point in time to the user;
enabling the user to access the point in time data; and
storing changes to the point in time in an undo stream.
7. The code of claim 4 further enabling:
determining to revert to the point in time; and
based on a positive determination stopping access to the second site, discarding the redo log on the first site, making the sites not fractured, synchronizing the sites to the point in time and allowing access to the second site.
8. The code of claim 7 further enabling:
based on a negative determination, stopping access to the first site discarding the changes made at point in time, rolling the first site to the latest point in time using the redo log, making the sites not fractured and synchronizing both sites to an image on the second site, and restoring access to the first site.
9. The code of claim 6 further enabling exposing the point in time to the user, where in the logical unit identity is different than the original logical unit identity.
10. The code of claim 9 further enabling: shutting down hosts on second site, before stopping access, and restarting hosts when access restored.
11. A computer implemented method use in replication, comprising:
quiescing IO commands to sites of an ACTIVE/ACTIVE storage system, the active/active storage system having at least two storage sites communicatively coupled via a virtualization layer; wherein the virtualization layer virtual service layer enables the at least two storage sites to present a same virtual volume at each of the at least two storage sites enabling transparent migration of a virtual machine between the at least two storage sites; wherein the virtualization layer enables simultaneous, active/active, concurrent read/write access to the same virtual volume at both of the sites; wherein the virtualization layer enables cache coherency between each of the at least two storage sited;
creating a change set for the same virtual volume;
unquiescing IO commands to the sites;
transferring data of a change set to the other sites of the active/active storage system by the virtualization layer, and
flushing the data; wherein flushing of the data of the change set enables the other sites of the at least two storage sites to synchronize changed data of the site from which the change set was transferred enabling the at least two storage sites to have a point in time with the same data.
12. The method of claim 11 further comprising:
reading, by the virtualization layer, the undo of the data to be flushed,
writing undo data corresponding to the data in the write cache to a data protection appliance (DPA);
writing, by the DPA, the undo data to a journal; and
writing the data to the backend storage system by the virtualization layer.
13. The method of claim 11 where the DPA is virtual machine and the method further comprising:
writing the data from the write cache to a volume;
notifying the DPA a full change set has been flushed; and
creating a bookmark at the undo journal by the DPA.
14. the code of claim 4 further comprising:
fracturing a cluster of an active/active storage system; wherein the cluster includes at least two sites; wherein an active/active storage system that is not fractured includes a virtualization layer enables the at least two storage sites to present a same virtual volume at each of the at least two storage sites enabling transparent migration of a virtual machine between the at least two storage sites; wherein the virtualization layer enables simultaneous, active/active, concurrent read/write access to the same virtual volume at both of the sites; wherein the virtualization layer enables cache coherency between each of the at least two storage sited; wherein a fractured active/active storage system enables an image on a first site of the sites to be rolled to a point in time and enables active production on a second site of the sites;
stopping IO on a first site of the cluster; and
rolling to a point in time on a first site of the cluster that is consistent with at least a second site of the cluster before the at least two sites were fractured.
15. The computer implemented method of claim 14 further comprising:
flushing the a write cache on a second site of the cluster to the first site of the cluster.
16. The computer implemented method of claim 14 further comprising:
requesting, by a user, access to a point in time;
writing undo information generated by rolling to the point in time to a redo log;
exposing the point in time to the user;
enabling the user to access the point in time data; and
storing changes to the point in time in an undo stream.
17. The computer implemented method of claim 14 further comprising:
determining to revert to the point in time; and
based on a positive determination stopping access to the second site, discarding the redo log on the first site, un-fracturing the sites by unfracturing the virtualization layer, synchronizing sites to the point in time and allowing access to the second site.
18. The computer implemented method of claim 17 further comprising:
based on a negative determination, stopping access to the first site discarding changes made at the point in time, rolling the first site to the latest point in time using the redo log, making the sites not fractured by connecting the virtualization layer and synchronizing both sites to an image on the second site, and restoring access to the first site.
19. The computer implemented method of claim 16 further comprising:
exposing the point in time to the user, where in the logical unit identity is different than the original logical unit identity.
20. The computer implemented method of claim 14 further comprising:
shutting down hosts on second site, before stopping access, and restarting hosts when access restored.