Synchronous replication of virtualized storage processors

Patent metadata

Field	Value
Publication number	US-9305009-B1
Application number	US-201314041191-A
Country	US
Kind code	B1
Filing date	Sep 30, 2013
Priority date	Sep 30, 2013
Publication date	Apr 5, 2016
Grant date	Apr 5, 2016

Abstract

A technique for managing replication of VSPs (Virtualized Storage Processors) proceeds on a per-VSP basis by (i) identifying the data objects associated with a VSP, (ii) establishing a common set of replication settings across all of the data objects associated with the VSP, and (iii) replicating the VSP by replicating each of the identified data objects associated with the VSP in accordance with the common set of replication settings established across all of the data objects associated with the VSP. The technique avoids the need for administrators to separately configure and manage replication on large numbers of individual data objects and thus reduces administrative burdens.

First claim

What is claimed is:

1. A method of managing synchronous replication of VSPs (Virtualized Storage Processors) in a data storage system, comprising:
establishing a first set of synchronous replication settings specific to a first VSP and a second set of synchronous replication settings specific to a second VSP, the first VSP including a first set of multiple data objects and the second VSP including a second set of multiple data objects distinct from the first set of multiple data objects, each data object of the first set of multiple data objects and the second set of multiple data objects being any of a file system, a LUN (logical storage unit), and a vVOL (virtual volume);
in response to the data storage system receiving first IO requests specifying data to be written to any of the first set of multiple data objects, synchronously replicating the first IO requests in accordance with the first set of synchronous replication settings specific to the first VSP; and
in response to the data storage system receiving second IO requests specifying data to be written to any of the second set of multiple data objects, synchronously replicating the second IO requests in accordance with the second set of synchronous replication settings specific to the second VSP.

2. The method of claim 1 , further comprising running a replication session manager to manage synchronous replication on a per-VSP basis.

3. The method of claim 2 , further comprising the replication session manager:
operating a synchronous replication session on each of the first set of multiple data objects in accordance with the first set of synchronous replication settings specific to the first VSP; and
operating a synchronous replication session on each of the second set of multiple data objects in accordance with the second set of synchronous replication settings specific to the second VSP.

4. The method of claim 3 , further comprising the replication session manager:
accessing a database that associates VSPs with data objects;
receiving from the database a first list of data objects associated with the first VSP, the first list of data objects identifying the first set of multiple data objects; and
receiving from the database a second list of data objects associated with the second VSP, the second list of data objects identifying the second set of multiple data objects.

5. The method of claim 3 , further comprising:
establishing a first set of recovery settings specific to the first VSP; and
establishing a second set of recovery settings specific to the second VSP.

6. The method of claim 4 , further comprising:
establishing a first set of failover settings specific to the first VSP; and
establishing a second set of failover settings specific to the second VSP.

7. The method of claim 6 , further comprising accepting user input for establishing the first and second sets of synchronous replication settings via a GUI (Graphical User Interface) application that serves as an entry point to the replication session manager.

8. The method of claim 3 , further comprising:
grouping the first set of multiple data objects into a first consistency group;
grouping the second set of multiple data objects into a second consistency group;
configuring a replication splitter in an IO stack of the data storage system to (i) respond to IO requests specifying data to be written to any data object in the first consistency group in accordance with the first set of synchronous replication settings and (ii) respond to IO requests specifying data to be written to any data object in the second consistency group in accordance with the second set of synchronous replication settings,
wherein synchronously replicating the first IO requests includes, for each of the first IO requests (i) the replication splitter receiving the first IO request, (ii) the replication splitter sending the first IO request to a replication appliance specified in the first set of synchronous replication settings, (iii) the replication splitter receiving an acknowledgement from the replication appliance in response to the first IO request to indicate that the first IO request has been persisted at a replication site, and (iv) processing the first IO request to effect a write operation on the data storage system, and
wherein synchronously replicating the second IO requests includes, for each of the second IO requests, (i) the replication splitter receiving the second IO request, (ii) the replication splitter sending the second IO request to a replication appliance specified in the second set of synchronous replication settings, (iii) the replication splitter receiving an acknowledgement from the replication appliance specified in the second set of synchronous replication settings in response to the second IO request to indicate that the second IO request has been persisted at a replication site, and (iv) processing the second IO request to effect a write operation on the data storage system.

9. The method of claim 8 , further comprising:
storing, in the data storage system, each of the first set of multiple data objects and each of the second set of multiple data objects in the form of a respective file that provides a realization of the respective data object; and
mapping each respective file to a respective logical volume;
wherein grouping the first set of multiple data objects into the first consistency group is effected by grouping into the first consistency group the local volumes to which the files that realize the first set of multiple data objects are mapped,
wherein grouping the second set of multiple data objects into the second consistency group is effected by grouping into the second consistency group the local volumes to which the files that realize the second set of multiple data objects are mapped, and
wherein the method further comprises the replication splitter (i) recognizing a logical volume to which each of the first IO requests and each of the second IO requests is directed, (ii) applying the first set of replication settings to IO requests directed to logical volumes that belong to the first consistency group, and (iii) applying the second set of replication settings to IO requests directed to logical volumes that belong to the second consistency group.

10. The method of claim 3 , further comprising the replication session manager establishing settings for performing synchronous replication of a VSP automatically and without user input for establishing the settings.

11. The method of claim 1 ,
wherein the first VSP further includes a collection of network servers that provide the first VSP with its own network identity, and
wherein the second VSP further includes a collection of network servers that provide the second VSP with its own network identity distinct from the network identity of the first VSP.

12. The method of claim 11 ,
wherein each of the first set of multiple data objects included in the first VSP is a file system, and
wherein each of the second set of multiple data objects included in the second VSP is a file system.

13. The method of claim 1 , further comprising:
grouping the first set of multiple data objects into a first consistency group;
grouping the second set of multiple data objects into a second consistency group; and
configuring an IO stack of the data storage system to (i) respond to IO requests specifying data to be written to any data object in the first consistency group in accordance with the first set of synchronous replication settings and (ii) respond to IO requests specifying data to be written to any data object in the second consistency group in accordance with the se

Assignees

• Emc Corp

Inventors

• Bono Jean-Pierre
• Tummala Himabindu
• Natanzon Assaf

Classifications

• G06F16/178 Primary

  Techniques for file synchronisation in file systems · CPC title

• G06F17/30174 Primary

  Physics · mapped topic