Migration of active virtual machines across multiple data centers

1. A computer-implemented method for performing multi-site migration of active virtual machines, comprising:

performing inter-site data migrations among a plurality of data centers separated by metropolitan-scale distances using a distributed volume system and a stretched cluster system to migrate the data synchronously within a distance and time latency limits defined by the distributed volume protocol;

wherein the distributed volume system comprises a clustered system using a distributed virtual volume as network storage media for the data migrations, wherein a cluster comprises a plurality of highly available and fully redundant enterprise-scale backup engines, each supporting on the order of thousands of virtual volumes that are accessible by all data centers of the plurality of data centers;

wherein the inter-site data migrations are performed by setting up distributed virtual volumes between pairs of data centers of the plurality of data centers to form migratable data clusters, and deploying pairs of migratable clusters as relay stations between each pair of distributed virtual volumes to link three data centers through the pair of distributed virtual volumes;

performing intra-site data migrations within each data center of the data centers using local live migration; and

combining the inter-site data migrations using the migratable cluster relay stations and the intra-site data migrations such that that data migrations among the plurality of data centers can be performed exclusively using synchronous transmission even if a distance between source and target data centers exceeds synchronous limits of the distributed volume protocol.

2. The method of claim 1 wherein the stretched volume system comprises a metro-scale virtualization system for storage clusters of the clustered system.

3. The method of claim 2 wherein the local live migration comprises a system encapsulating an entire state of a virtual machine by a set of files stored on the network storage media and transferring an execution state of the virtual machine rapidly from a source computer server to a destination server computer.

4. The method of claim 1 wherein the distributed virtual volumes between pairs of data centers of the plurality of data centers effectively form a single distributed virtual volume accessible by all data centers of the plurality of data centers.

5. The method of claim 1 wherein the synchronous limits comprise at least one of a distance exceeding 100 kilometers, or a round-trip time of data transfer in excess of 10 milliseconds.

6. The method of claim 1 further comprising performing at least one routing optimization, load balancing, or least distance algorithm to optimize the combining of the inter-site and intra-site data migrations.

7. A system configured to perform multi-site migration of active virtual machines in a large-scale network, comprising:

a plurality of physical storage devices configured to store data of the active virtual machines during the multi-site migration;

a backup server computer coupled to the physical storage devices performing inter-site data migrations among a plurality of data centers separated by metropolitan-scale distances using a distributed volume system and a stretched cluster system to migrate the data synchronously within a distance and time latency limits defined by the distributed volume protocol;

wherein the distributed volume system comprises a clustered system using a distributed virtual volume as network storage media for the data migrations, wherein a cluster comprises a plurality of highly available and fully redundant enterprise-scale backup engines, each supporting on the order of thousands of virtual volumes that are accessible by all data centers of the plurality of data centers; and

wherein the inter-site data migrations are performed by setting up distributed virtual volumes between pairs of data centers of the plurality of data centers to form migratable data clusters, and deploying pairs of migratable clusters as relay stations between each pair of distributed virtual volumes to link three data centers through the pair of distributed virtual volumes; and

a backup component performing intra-site data migrations within each data center of the data centers using local live migration, and combining the inter-site data migrations using the migratable cluster relay stations and the intra-site data migrations such that that data migrations among the plurality of data centers can be performed exclusively using synchronous transmission even if a distance between source and target data centers exceeds synchronous limits of the distributed volume protocol.

8. The system of claim 7 wherein the stretched volume system comprises a metro-scale virtualization system for storage clusters of the clustered system.

9. The system of claim 8 wherein the local live migration comprises a system encapsulating an entire state of a virtual machine by a set of files stored on the network storage media and transferring an execution state of the virtual machine rapidly from a source computer server to a destination server computer.

10. The system of claim 7 wherein the distributed virtual volumes between pairs of data centers of the plurality of data centers effectively form a single distributed virtual volume accessible by all data centers of the plurality of data centers.

11. The system of claim 7 wherein the synchronous limits comprise at least one of a distance exceeding 100 kilometers, or a round-trip time of data transfer in excess of 10 milliseconds.

12. The system of claim 7 further comprising an optimization component performing at least one routing optimization, load balancing, or least distance algorithm to optimize the combining of the inter-site and intra-site data migrations.

13. A computer program product, comprising a non-transitory computer-readable medium having a computer-readable program code embodied therein, the computer-readable program code adapted to be executed by one or more processors to perform multi-site migration of active virtual machines by executing instructions implementing a method comprising:

performing inter-site data migrations among a plurality of data centers separated by metropolitan-scale distances using a distributed volume system and a stretched cluster system to migrate the data synchronously within the distance and time latency limits defined by the distributed volume protocol;

wherein the distributed volume system comprises a clustered system using a distributed virtual volume as network storage media for the data migrations, wherein a cluster comprises a plurality of highly available and fully redundant enterprise-scale backup engines, each supporting on the order of thousands of virtual volumes that are accessible by all data centers of the plurality of data centers; and

wherein the inter-site data migrations are performed by setting up distributed virtual volumes between pairs of data centers of the plurality of data centers to form migratable data clusters, and deploying pairs of migratable clusters as relay stations between each pair of distributed virtual volumes to link three data centers through the pair of distributed virtual volumes; and

performing intra-site data migrations within each data center of the data centers using local live migration; and

combining the inter-site data migrations using the migratable cluster relay stations and the intra-site data migrations such that that data migrations among the plurality of data centers can be performed exclusively using synchronous transmission even if a distance between source and target data centers exceeds synchronous limits of the distributed volume protocol.

14. The computer-program product of claim 13 wherein the method further comprises performing at least one routing optimization, load balancing, or least distance algorithm to optimize the combining of the inter-site and intra-site data migrations.