Performing pre-stage replication of data associated with virtual machines prior to migration of virtual machines based on resource usage

摘要

Embodiments of the present invention relate to the migration of virtual machines (VMs) between networked computing environments (e.g., cloud computing environments) based on resource utilization. Specifically, embodiments of the present invention provide an approach to select an optimal set (one or more) of VMs as candidates for pre-staged migration. In a typical embodiment, when a first cloud environment nears physical resource capacity, an optimal set of VMs will be identified for migration to a second cloud environment that has sufficient capacity to accommodate workload(s) from the first cloud environment. To make this process more efficient, data associated with the set of virtual machines may be “pre-stage” replicated from the first cloud environment to the second cloud environment (e.g., in advance of the migration of the identified set of VMs).

主权项

1. A computer-implemented method for migrating virtual machines between networked computing environments, comprising:
determining that resource usage of resources of a first networked computing environment by the first networked computing environment is approaching a predetermined level, wherein a determination that the predetermined level is being approached is made when monitoring of the resource usage over time shows a trend that indicates that the predetermined level will be reached within a predetermined period of time; identifying a set of virtual machines in the first networked computing environment as a candidate for migration to a second networked computing environment, each virtual machine in the set of virtual machines being identified based upon having at least one of the following criteria relative to non-candidates: a smaller size of data associated with the virtual machine, a lower frequency of change of the data associated with the virtual machine, a longer lifecycle of the virtual machine, or a more predictable pattern level of pattern predictability associated with usage of the virtual machine; performing, in response to a determination that the predetermined level is being approached, a pre-stage replication of the data associated with the set of virtual machines identified as the candidate to the second networked computing environment; operating, subsequent to the pre-stage replication in response to the determination that the predetermined level is being approached, the set of virtual machines within the first networked computing environment so long as the predetermined level has not been reached; migrating, in response to a determination that the predetermined level has been reached, at least one virtual machine of the set of virtual machines from the first networked computing environment to the second networked computing environment; and associating, responsive to the migrating, the data with the at least one virtual machine in the second networked computing environment.