Leveraging array snapshots for immediate continuous data protection
One challenge to continuous data protection (CDP) is the resources required for storage. Typically, when setting up replication, a production volume and replica are set up and initialized to copy data from the production volume to the replica. However, instead of taking a clone of the volume, there are significant advantages to taking an array-based snapshot of the volume and leveraging capabilities inherent to the snapshot to improve data protections systems using CDP. To leverage the benefits of snapshots in a CDP system, as described below, example embodiments of the present invention use multiple snapshots. Bookmarks are created to indicate the point in time each snapshot was create so that a newer snapshot may be used for replication as data is applied to the replica.
1. A method for leveraging snapshots for immediate continuous data protection (CDP) in a data replication environment comprising a production site comprising a host, a storage array, and a splitter, the method comprising:
creating a replica of a production volume at the production site;
configuring a replication service to replicate from the production volume to the replica of the production volume; and
performing replication to the replica of the production volume using changes tracked at the production site, comprising:
attaching the production volume to the splitter as clean;
attaching the replica to the splitter as clean;
tracking input/output (I/O) commands arriving at the splitter; and
starting replication to the replica of the production volume using the tracked I/O commands.
2. The method of claim 1 wherein the replica is a clone or a snapshot.
3. The method of claim 1 wherein the production volume is among a plurality of volumes in a consistency group and wherein creating a replica of a production volume comprises creating respective replicas of all volumes in the consistency group contemporaneously.
4. The method of claim 1 wherein the production volume is among a plurality of volumes in a consistency group and wherein creating a replica of the production volume comprises:
creating a respective replica of each volume in the consistency group sequentially; and
making the replicas consistent at the same point in time by applying changes from a journal to each replica.
5. The method of claim 1 wherein the replica of the production volume is a first snapshot, the method further comprising:
creating a first bookmark in a journal;
creating a second snapshot of the production volume; and
creating a second bookmark in the journal.
6. The method of claim 5 further comprising maintaining thin snapshots by updating the snapshot of the production volume with journal data.
7. The method of claim 5 wherein updating the snapshot of the production volume comprises:
reading the first bookmark from a do stream of the journal; and
applying replication data to the first snapshot and the second snapshot.
8. The method of claim 7 wherein updating the snapshot of the production volume further comprises:
reading the second bookmark from the do stream of the journal;
deleting the first snapshot;
applying the replication data to the second snapshot; and
using the second snapshot as the replica copy.
9. An apparatus for leveraging snapshots for immediate continuous data protection (CDP), the apparatus comprising:
a replicator configured to create a replica of a production volume at a production site; and
a manager configured to configure a replication service to replicate from the production volume to the replica of the production volume at the production site, perform replication using changes tracked at the production site, attach the production volume to a splitter as clean, attach the replica to the splitter as clean, track input/output (I/O) commands arriving at the splitter and start replication to the replica of the production volume using the tracked I/O commands.
10. The apparatus of claim 9 wherein the replica is a clone or a snapshot.
11. The apparatus of claim 9 wherein the production volume is among a plurality of volumes in a consistency group and wherein the replicator is further configured to create respective replicas of all volumes in the consistency group contemporaneously.
12. The apparatus of claim 9 wherein the production volume is among a plurality of volumes in a consistency group and wherein the replicator is further configured to create respective replicas of each volume in the consistency group sequentially and make the replicas consistent at the same point in time by applying changes from a journal to each replica.
13. The apparatus of claim 9 wherein the replica of the production volume is a first snapshot and wherein the replicator is further configured to create a first bookmark in a journal, create a second snapshot of the production volume and create a second bookmark in the journal.
14. The apparatus of claim 13 further comprising a journal configured to provide a protection window and wherein the manager is further configured to maintain thin snapshots by updating the snapshot of the production volume with journal data.
15. The apparatus of claim 13 wherein the manager is further configured to read the first bookmark from a do stream of the journal and apply replication data to the first snapshot and the second snapshot.
16. The apparatus of claim 15 wherein the manager is further configured to read the second bookmark from the do stream of the journal, delete the first snapshot, apply replication data to the second snapshot and use the second snapshot as the replica copy.
17. A computer-program product including a non-transitory computer-readable storage medium encoded with computer-program code that, when executed on a processor of a computer, cause the computer to leverage snapshots for immediate continuous data protection (CDP), the computer-program code comprising:
computer-program code for creating a replica of a production volume at a production site;
computer-program code for configuring a replication service to replicate from the production volume to the replica of the production volume at the production site; and
computer-program code for performing replication using changes tracked at the production site, comprising:
computer-program code for attaching the production volume to a splitter as clean;
computer-program code for attaching the replica to the splitter as clean;
computer-program code for tracking input/output (I/O) commands arriving at the splitter; and
computer-program code for starting replication to the replica of the production volume using the tracked I/O commands.
18. The computer-program product of claim 17 wherein the replica is a clone or a snapshot.
19. The computer-program product of claim 17 wherein the production volume is among a plurality of volumes in a consistency group and wherein computer-program code for creating a replica of a production volume comprises computer-program code for creating respective clones of all volumes in the consistency group contemporaneously or creating a respective clone of each volume in the consistency group sequentially and computer-program code for making the clones consistent at the same point in time by applying changes from a journal to each clone.
20. The computer-program product of claim 17 wherein a journal provides a protection window and wherein the replica of the production volume is a first snapshot, the computer-program product further comprising:
computer-program code for creating a first bookmark in the journal;
computer-program code for creating a second snapshot of the production volume;
computer-program code for creating a second bookmark in the journal; and
computer-program code for maintaining thin snapshots by updating the snapshot of the production volume with journal data.
21. The computer-program product of claim 20 wherein computer-program code for updating the snapshot of the production volume comprises:
computer-program code for reading the first bookmark from a do stream of the journal;
computer-program code for applying replication data to the first snapshot and the second snapshot;
computer-program code for reading the second bookmark from the do stream of the journal;
computer-program code for deleting the first snapshot;
computer-program code for applying the replication data to the second snapshot; and
computer-program code for using the second snapshot as the replica copy.