Recovery of a logical unit in a consistency group while replicating other logical units in the consistency group

• US 8,954,796 B1
• Filed: 06/26/2012
• Issued: 02/10/2015
• Est. Priority Date: 06/26/2012
• Status: Active Grant

• Current Assignee: EMC International Company (Dell Technologies Inc.)
• Original Assignee: EMC International Company (Dell Technologies Inc.)
• Inventors: Natanzon, Assaf, Cohen, Saar
• Primary Examiner(s): Duncan, Marc
• Application Number:
  US13/533,137
• Time in Patent Office: 959 Days
• Field of Search: 714/15, 714/16, 714/19, 714/20, 707/674, 707/683
• US Class Current: 714/15
• CPC Class Codes: G06F 11/1471 involving logging of persistent data for recovery
• CPC Class Codes: G06F 11/2064 while ensuring consistency
• CPC Class Codes: G06F 11/2074 Asynchronous techniques
• CPC Class Codes: G06F 2201/815 Virtual
• CPC Class Codes: G06F 2201/855 Details of asynchronous mirroring using a journal to transfer not-yet-mirrored changes

Abstract

In one aspect, a method includes recovering at least one logical unit of a consistency group in a journal-based replication system without interruption of replication of other logical units in the consistency group, the at least one logical unit being one of a physical logical unit or a virtual logical unit.

Claims

1. A method comprising:

recovering at least one logical unit of a consistency group in a journal-based replication system without interruption of replication of other logical units in the consistency group, the at least one logical unit being one of a physical logical unit or a virtual logical unit;

receiving a request to recover a logical unit in the consistency group at a point-in-time (PIT);

ceasing replication of the logical unit to a backup logical unit;

recovering the logical unit at the PIT while continuing replicating other logical units in the CG;

marking a first recovery point at an end portion of an UNDO METADATA stream associated with the PIT;

marking a second recovery point at an end portion of a DO METADATA stream associated with the PIT; and

generating differences between the production logical unit and the backup logical unit by joining the metadata in a local delta marker stream and the data in the DO metadata stream associated with the production logical unit.

2. The method of claim 1, further comprising copying the differences from the backup logical unit to the production logical unit.

3. The method of claim 2, further comprising, after the production logical unit is recovered:

accepting new I/O requests; and

replicating the new I/Os.

4. A method comprising:

recovering at least one logical unit of a consistency group in a journal-based replication system without interruption of replication of other logical units in the consistency group, the at least one logical unit being one of a physical logical unit or a virtual logical unit;

receiving a request to recover a logical unit in the consistency group at a point-in-time (PIT);

ceasing replication of the logical unit to a backup logical unit;

recovering the logical unit at the PIT while continuing replicating other logical units in the CG;

marking a first recovery point at an end portion of an UNDO METADATA stream associated with the PIT;

marking a second recovery point at an end portion of a DO METADATA stream associated with the PIT;

reading data from a write transaction entry in a DO stream;

determining if the write transaction entry is between the first recovery point and the second recovery point; and

discarding the write transaction entry if the write transaction entry is between the first recovery point and the second recovery point; and

if the write transaction entry is not between the first recovery point and the second recovery point;

reading data from an UNDO stream for the backup logical unit;

writing the UNDO data read from the logical unit to the UNDO stream of the production logical unit; and

applying the write transaction read from the DO stream to the backup logical unit, if the entry is to a device different than the recovered device.

5. An article comprising:

a non-transitory machine-readable medium that stores executable instructions, the instructions causing a machine to;

recover at least one logical unit of a consistency group in a journal-based replication system without interruption of replication of other logical units in the consistency group, the at least one logical unit being one of a physical logical unit or a virtual logical unit;

receive a request to recover a logical unit in the consistency group at a point-in-time (PIT);

cease replication of the logical unit to a backup logical unit;

recover the logical unit at the PIT while continuing replicating other logical units in the CG;

mark a first recovery point at an end portion of an UNDO METADATA stream associated with the PIT;

mark a second recovery point at an end portion of a DO METADATA stream associated with the PIT; and

generate differences between the production logical unit and the backup logical unit by joining the metadata in a local delta marker stream and the data in the DO metadata stream associated with the production logical unit.

6. The article of claim 5, further comprising instructions causing the machine to:

read data from a write transaction entry in a DO stream;

determine if the write transaction entry is between the first recovery point and the second recovery point; and

discard the write transaction entry if the write transaction entry is between the first recovery point and the second recovery point; and

if the write transaction entry is not between the first recovery point and the second recovery point;

read data from an UNDO stream for the backup logical unit;

write the UNDO data read from the logical unit to the UNDO stream of the production logical unit; and

apply the write transaction read from the DO stream to the backup logical unit, if the entry is to a device different than the recovered device.

7. The article of claim 5, further comprising instructions causing the machine to:

copy the differences from the backup logical unit to the production logical unit; and

after the production logical unit is recovered;

accept new I/O requests; and

replicate the new I/Os.

8. An apparatus, comprising:

circuitry configured to;

recover at least one logical unit of a consistency group in a journal-based replication system without interruption of replication of other logical units in the consistency group, the at least one logical unit being one of a physical logical unit or a virtual logical unit;

receive a request to recover a logical unit in the consistency group at a point-in-time (PIT);

cease replication of the logical unit to a backup logical unit; and

recover the logical unit at the PIT while continuing replicating other logical units in the CG;

mark a first recovery point at an end portion of an UNDO METADATA stream associated with the PIT;

mark a second recovery point at an end portion of a DO METADATA stream associated with the PIT; and

generate differences between the production logical unit and the backup logical unit by joining the metadata in a local delta marker stream and the data in the DO metadata stream associated with the production logical unit.

9. The apparatus of claim 8 wherein the circuitry comprises at least one of a processor, a memory, programmable logic and logic gates.

10. The apparatus of claim 8, wherein the circuitry is further configured to:

read data from a write transaction entry in a DO stream;

determine if the write transaction entry is between the first recovery point and the second recovery point; and

discard the write transaction entry if the write transaction entry is between the first recovery point and the second recovery point; and

if the write transaction entry is not between the first recovery point and the second recovery point;

read data from an UNDO stream for the backup logical unit;

write the UNDO data read from the logical unit to the UNDO stream of the production logical unit; and

apply the write transaction read from the DO stream to the backup logical unit, if the entry is to a device different than the recovered device.

11. The apparatus of claim 8, wherein the circuitry is further configured to:

copy the differences from the backup logical unit to the production logical unit; and

after the production logical unit is recovered;

accept new I/O requests; and

replicate the new I/Os.