Service migration across cluster boundaries

摘要

Embodiments provide migration of services across different clusters to balance utilization and meet customer demands. Different service migration options may be performed with or without downtime. The artifacts of the service are moved to a new destination cluster. The service is created on the new destination cluster and staged so that the service is almost ready to start. In one embodiment, the service is stopped on the old cluster and started on the new cluster. After stopping the service, DNS is updated to point to the service on the new cluster. In another embodiment, the service is stopped on the old cluster and started on the new cluster with the same IP address to avoid DNS reprogramming and associated delays. In a further embodiment, the migration is performed without downtime by moving the service part by part from one cluster to another.

主权项

1. A computer-implemented method, comprising:
copying artifacts from a first instance of a service running on a first cluster in a computing environment on a network, the first cluster comprising a first group of virtual machines to run the first instance of the service, the first instance of the service comprising a first plurality of parts in a running state and being associated with an IP address; creating a second instance of the service on a second cluster in the computing environment on the network by selecting one or more nodes on the second cluster and configuring a second group of virtual machines to run the second instance of the service thereon using the artifacts copied from the first instance of the service, the second instance of the service comprising a second plurality of parts in a non-running state, each of the second plurality of parts corresponding to one of the first plurality of parts and also being associated with the IP address; stopping, on the first instance of the service on the first cluster, a first part from the first plurality of parts so that the first part from the first plurality of parts is in the non-running state; starting, on the second instance of the service on the second cluster, a corresponding part from the second plurality of parts so that the corresponding part from the second plurality of parts is in the running state; updating the network to provide access to only the parts from the first and second plurality of parts that are in the running state; performing the stopping, starting, and updating steps two or more times until all parts from the first plurality of parts is in the non-running state and all corresponding parts from the second plurality of parts is in the running state; and deleting the first instance of the service on the first cluster.