| 主权项 |
1. A method for virtual machine load balancing, comprising:
monitoring, from a host hardware platform, specification and utilization information for a plurality of peer hardware platforms networked together on a network, wherein each of the plurality of peer hardware platforms is configured to publish its own specification and utilization information on the network, and wherein the peer hardware platforms are each configured to run virtual machines; accessing current consumption data and historical consumption data for a first virtual machine running on the host hardware platform, wherein the host hardware platform is coupled to the network; and determining that the first virtual machine running on the host hardware platform is over-constrained; selecting, by the host hardware platform, a target platform from among the peer hardware platforms to receive the first virtual machine, based on the published specification and utilization information for the peer hardware platforms and consumption data for the first virtual machine, wherein the first virtual machine will operate on the target platform in a condition that is not over-constrained; wherein each peer hardware platform is configured to select for itself another peer hardware platform, from among the other peer hardware platforms, to receive an over-constrained virtual machine currently running on the peer hardware platform making the selection, wherein the selection is made based on the published specification and utilization information for the peer hardware platforms and consumption data for the over-constrained virtual machine. |
