Mirror vote synchronization

Patent metadata

Field	Value
Publication number	US-9361194-B2
Application number	US-201414261737-A
Country	US
Kind code	B2
Filing date	Apr 25, 2014
Priority date	Mar 20, 2014
Publication date	Jun 7, 2016
Grant date	Jun 7, 2016

Abstract

One or more techniques and/or systems are provided for mirror vote synchronization. For example, a first storage device is located at a first storage site, and a second storage device is located at a second storage site. The second storage device is configured according to a data mirroring configuration where data from the first storage device is mirrored to the second storage device. Mirror vote metadata is generated based upon an up-to-date state of the data mirroring configuration. The mirror vote metadata indicates whether the first storage device and/or the second storage device are up-to-date or not. The mirror vote metadata may be replicated between the first storage site and the second storage site. If the first storage site fails, then the second storage site may provide switchover operation using the second storage device based upon the mirror vote metadata.

First claim

What is claimed is:

1. A method, comprising:
identifying, by a storage server, a first storage device maintained at a first storage site;
identifying a second storage device maintained at a second storage site, the second storage device configured according to a data mirroring configuration for the first storage device;
generating a local mirror vote based upon an up-to-date state of the data mirroring configuration;
replicating the local mirror vote from the first storage site to the second storage site to create a replicated mirror vote; and
responsive to identifying a disaster of the first storage site:
assigning ownership of the second storage device to the second storage site as a switchover storage aggregate;
responsive to the replicated mirror vote indicating that the second storage device is up-to-date, placing the switchover storage aggregate into an operational state; and
responsive to the replicated mirror vote indicating that the second storage device is out-of-date, placing the switchover storage aggregate into a non-operational state.

2. The method of claim 1 , the generating a local mirror vote comprising at least one of:
storing the local mirror vote within a first root storage location of the first storage site; and
storing the local mirror vote within a first replication database.

3. The method of claim 1 , comprising:
storing the replicated mirror vote within a second root storage location of the second storage site.

4. The method of claim 1 , comprising:
facilitating data access to the switchover storage aggregate utilizing the second storage device; and
generating a switchover mirror vote based upon the data access, the switchover mirror vote indicating that the first storage device is out-of-date and that the second storage device is up-to-date.

5. The method of claim 4 , the generating a switchover mirror vote comprising at least one of:
storing the switchover mirror vote within a second root storage location of the second storage site; and
storing the switchover mirror vote within a second replication database.

6. The method of claim 4 , comprising:
responsive to identifying a restoration of the first storage site:
replicating the switchover mirror vote from the second storage site to the first storage site as a restored mirror vote; and
assigning ownership of the second storage device to the first storage site.

7. The method of claim 6 , comprising:
facilitating selective operation of the second storage device based upon the restored mirror vote, the facilitating comprising:
evaluating the restored mirror vote with an up-to-date mirror vote; and
responsive to the restored mirror vote being different than the up-to-date mirror vote, selecting either the restored mirror vote or the up-to-date mirror vote for selective operation of the first storage device and for selective operation of the second storage device based upon a tie-breaker technique.

8. The method of claim 4 , comprising:
responsive to identifying a restoration of the first storage site:
assigning ownership of the first storage device to the second storage site;
performing a synchronization between the second storage device and the first storage device based upon the switchover mirror vote;
generating a synchronized mirror vote based upon the synchronization;
providing the synchronized mirror vote to the first storage site; and
assigning ownership of the first storage device and the second storage device to the first storage site for selective operation of the first storage device and the second storage device based upon the synchronized mirror vote.

9. The method of claim 4 , comprising:
responsive to identifying a restoration of the first storage site:
assigning ownership of the first storage device to the second storage site;
performing a synchronization between the second storage device and the first storage device based upon the switchover mirror vote; and
responsive to completion of the synchronization:
identifying a failure associated with the second storage device during the synchronization;
generating a heal failure mirror vote based upon the failure, the heal failure mirror indicating that the first storage device is up-to-date and that the second storage device is out-of-date; and
responsive to identifying restoration of the second storage device:
providing the heal failure mirror vote to the first storage site; and
assigning ownership of the first storage device and the second storage device to the first storage site for selective operation of the first storage device and the second storage device based upon the heal failure mirror vote.

10. The method of claim 1 , the assigning ownership comprising:
initializing the switchover storage aggregate, the initializing comprising:
responsive to determining that a second replication database is not online:
initializing the switchover storage aggregate in an offline state; and
setting a flag for the switchover storage aggregate, the flag specifying that the switchover storage aggregate is in the offline state because mirror vote metadata was unavailable for the switchover storage aggregate during the initializing.

11. The method of claim 10 , comprising:
responsive to determining that the second replication database is online:
removing the flag from the switchover storage aggregate; and
determining whether to place the switchover first storage aggregate into the operational state or the non-operational state based upon the replicated mirror vote.

12. The method of claim 1 , the local mirror vote stored within a first replication database associated with the first storage site and the replicated mirror vote stored within a second replication database associated with the second storage site.

13. A computing device comprising:
a memory containing machine readable medium comprising machine executable code having stored thereon instructions for performing a method; and
a processor coupled to the memory, the processor configured to execute the machine executable code to cause the processor to:
identify a first storage device at a first storage site;
identify a second storage device at a second storage, the second storage device configured according to a data mirroring configuration for the first storage device;
generate a local mirror vote on the first storage site based upon an up-to-date state of the data mirroring configuration;
replicate the local mirror vote from the first storage site to the second storage site to create a replicated mirror vote; and
responsive to identifying a disaster of the first storage site:
expose the replicated mirror vote through the second storage site, the second storage site assigned ownership of the second storage device as a switchover storage aggregate, the replicated mirror vote indicating whether the second storage site is to place the switchover storage aggregate into an operational state or a non-operational state.

14. The computing device of claim 13 , the local mirror vote stored within a first replication database associated with the first storage site and the replicated mirror vote stored within a second replication database associated with the second storage site.

15. The computing device of claim 13 , wherein the machine executable code causes the processor to at least one of:
store the local mirror vote within a first root storage location of the first storage site; and
store the local mirror vote within a first replication database.

16. The computing device of claim 13 , wherein the machine executable code causes the

Assignees

• Netapp Inc

Inventors

• Long Brandon Taylor
• Riedle Linda Ann
• Kulkami Manali
• Nirmale Sandeep T
• Krishnamurthy Vikram Harakere

Classifications

• G06F11/2069 Primary

  Management of state, configuration or failover · CPC title

• G06F11/2082

  Data synchronisation · CPC title

• G06F11/2071

  using a plurality of controllers · CPC title

• H04L67/1095

  Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes · CPC title

• G06F3/065 Primary

  Replication mechanisms · CPC title