| 主权项 |
1. A method for handling failure in a networked virtualization environment having distributed virtual machine management, comprising:
identifying a failed management virtual machine instance of a plurality of management virtual machine instances in the networked virtualization environment,
wherein the networked virtualization environment comprises a plurality of nodes, where a node of the plurality of nodes comprises a hypervisor, one or more virtualization components and a management virtual machine instance that runs on top of the hypervisor,wherein each management virtual machine instance of the plurality of management virtual machine instances services a subset of virtualization components in the networked virtualization environment, andwherein a management virtual machine instance of the plurality of management virtual machine instances has access to a corresponding shard of a distributed database for the networked virtualization environment, the corresponding shard storing a subset of management data for virtualization components managed by the management virtual machine instance; designating one or more active management virtual machine instances of the plurality of management virtual machine instances for replacing virtualization component management functionality of the failed management virtual machine instance,
wherein the one or more active management virtual machine instances replace virtualization component management functionality of the failed management virtual machine instance by each accessing its subset of management data in its corresponding shard of the distributed database, andwherein the one or more active management virtual machine instances are configured to replace virtualization component management functionality of another management virtual machine instance whenever the other management virtual machine instance fails; and distributing a workload of the failed management virtual machine instance amongst the one or more active management virtual machine instances in response to identifying the failed management virtual machine instance. |
