AntiAfinity

Patent metadata

Field	Value
Publication number	US-9619264-B1
Application number	US-201213599235-A
Country	US
Kind code	B1
Filing date	Aug 30, 2012
Priority date	Jun 27, 2012
Publication date	Apr 11, 2017
Grant date	Apr 11, 2017

Abstract

A computer implemented method, system, and computer program product for recovering from a crash of a system being replicated, the method comprising determining the amount of recovery time due to the crash of each of a set of hypervisors; wherein each of the hypervisors runs one or more data replication elements selected from the group consisting of a splitter and a replication appliance; wherein each of the splitters and replication appliances replicates one or more volumes, creating an assignment of the one or more volumes to the set of replication appliances and creating an assignment of one or more replication appliances to a set of hypervisors to minimize the amount of recovery time.

First claim

What is claimed is:

1. A computer implemented method for recovering from a crash of a system being replicated, the method comprising:
determining the amount of recovery time for a set of consistency groups due to a crash of each of a set of hypervisors; wherein the recovery time corresponds to amount of time to resynchronize data being replicated by the consistency group due to the crash of each of the set of hypervisors; wherein each of the hypervisors runs one or more data replication elements selected from a group consisting of a splitter and a replication appliance for the set of consistency group; wherein each of the splitters and replication appliances replicates one or more volumes of the set of consistency groups containing one or more replication sets; wherein the splitter intercepts IO written to one or more volumes and replicates a copy of the intercepted IO; wherein a volume running in a crashed hypervisor being replicated before the crash to a data replication appliance in the crashed hypervisor needs to be resynchronized; wherein a volume running in the crashed hypervisor being replicated to a data replication appliance in a different hypervisor does not need to be resynchronized;
creating an assignment of the one or more volumes to the replication appliances for the set of consistency groups and creating an assignment of one or more replication appliances to the set of hypervisors for the set of consistency groups to minimize an amount of recovery time by minimizing the amount of data needed to be resynchronized for each consistency group; wherein the assignment of the one or more volumes to the set of replication appliances denotes which replication appliance replicates which of the one or more volumes; wherein the assignment of one or more replication appliances to the set of hypervisors denotes which replication appliance runs on which hypervisor.

2. The computer implemented method of claim 1 further comprising:
moving one or more of the volumes to another replication appliance running on another hypervisor to minimize recovery time in the event of a crash of one of the hypervisors of the set of hypervisors; wherein moving the volume to a different replication appliance lowers the amount of recovery time for the hypervisor running the replication appliance.

3. The computer implemented method of claim 2 wherein the volumes are assigned to one or more consistency groups and wherein the moving comprises:
moving one or more consistency groups from a replication appliance running on a first hypervisor to a second replication appliance running on a second hypervisor to minimize recovery time in the event of a crash of one of the hypervisors of the set of hypervisors; wherein moving the consistency group to a different replication appliance lowers the amount of recovery time for the hypervisor running the replication appliance.

4. The computer implemented method of claim 1 wherein the creating comprises:
using an optimization technique selected from a group consisting of simulated annealing, genetic algorithms, genetic programming, and neural networks to create the assignment.

5. The computer implemented method of claim 4 wherein the determining the amount of recovery time comprises:
finding for each replication appliance, for each hypervisor of the set of hypervisors, the load on the replication appliance wherein the load include the amount of data being replicated by the replication appliance.

6. The computer implemented method of claim 4 wherein the determining the amount of recovery time comprises:
determining, for each hypervisor of the set of hypervisors, the amount of data replicated by the splitter and an appliance running on the hypervisor; wherein the amount of data replicated by a splitter corresponds to the amount of IO intercepted and split by the splitter.

7. The computer implemented method of claim 1 further comprising moving a virtual appliance from one hypervisor to another hypervisor using vmotion; wherein the virtual appliance is running in a virtual machine and the vmotion moves the virtual machine running the virtual appliance.

8. A computer program product for recovering from a crash of a system being replicated comprising:
a non-transitory computer readable medium encoded with computer executable program code for replication of data, the computer executable program code executed across one or more processors to perform:
determining the amount of recovery time for a set of consistency groups due to a crash of each of a set of hypervisors; wherein the recovery time corresponds to amount of time to resynchronize data being replicated by the consistency group due to the crash of each of the set of hypervisors; wherein each of the hypervisors runs one or more data replication elements selected from a group consisting of a splitter and a replication appliance for the set of consistency group; wherein each of the splitters and replication appliances replicates one or more volumes of the set of consistency groups containing one or more replication sets; wherein the splitter intercepts IO written to one or more volumes and replicates a copy of the intercepted IO; wherein a volume running in a crashed hypervisor being replicated before the crash to a data replication appliance in the crashed hypervisor needs to be resynchronized; wherein a volume running in the crashed hypervisor being replicated to a data replication appliance in a different hypervisor does not need to be resynchronized;
creating an assignment of the one or more volumes to the replication appliances for the set of consistency groups and creating an assignment of one or more replication appliances to the set of hypervisors for the set of consistency groups to minimize the amount of recovery time by minimizing the amount of data needed to be resynchronized for each consistency group; wherein the assignment of the one or more volumes to the set of replication appliances denotes which replication appliance replicates which of the one or more volumes; wherein the assignment of one or more replication appliances to the set of hypervisors denotes which replication appliance runs on which hypervisor.

9. The computer program product of claim 8 the computer executable program code further executed to perform:
moving one or more of the volumes to a different replication appliance running on a different hypervisor to minimize recovery time in the event of a crash of one of the hypervisors of the currently amended of hypervisors; wherein moving the volume to a different replication appliance lowers the amount of recovery time for the hypervisor running the replication appliance.

10. The computer program product of claim 9 wherein the volumes are assigned to one or more consistency groups and the computer executable program code further executed to perform:
moving one or more consistency groups from a replication appliance running on a first hypervisor to a second replication appliance running on a second hypervisor to minimize recovery time in the event of a crash of one of the hypervisors of the set of hypervisors wherein moving the consistency group to a different replication appliance lowers the amount of recovery time for the hypervisor running the replication appliance.

11. The computer program product of claim 8 wherein the computer executable program code further executed to perform:
using an optimization technique selected from a group consisting of simulated annealing, genetic algorithms, genetic programming, and neural networks to create the assignment.

12. The computer program product of claim 11 the computer program product code further executed to perform:
finding each re

Assignees

• Natanzon Assaf
• Cohen Saar
• Shemer Jehuda
• Solan Alex
• Emc Ip Holding Co Llc

Inventors

• Natanzon Assaf
• Cohen Saar
• Shemer Jehuda
• Solan Alex

Classifications

• G06F9/45558 Primary

  Hypervisor-specific management and integration aspects · CPC title

• G06F2009/45591

  Monitoring or debugging support · CPC title

• G06F11/2048

  where the redundant components share neither address space nor persistent storage · CPC title

• G06F11/2094

  Redundant storage or storage space (G06F11/2056 takes precedence) · CPC title

• G06F11/2097

  maintaining the standby controller/processing unit updated (initialisation or re-synchronisation thereof G06F11/1658 and subgroups) · CPC title