Does Kubernetes provision new VMs for pods on my cloud platform? -
i'm learning kubernetes , still trying figure out. general use of think there still plenty of things i'm missing, here's 1 of them. if want run kubernetes on public cloud, gce or aws, kubernetes spin new vms in order make more compute new pods might needed? or use amount of vms pre-configured compute pool. heard brendan say, in talk in coreos fest, kubernetes sees vms "sea of compute" , user doesn't have worry vm running pod - i'm interested know pool of compute comes from, configured when setting kubernetes? or scale , create new machines needed?
i hope managed coherent. thanks!
kubernetes supports scaling, not auto-scaling. addition , removal of new pods (vms) in kubernetes cluster performed replication controllers. size of replication controller can changed updating replicas
field. can performed in couple ways:
- using kubectl, can use scale command.
- using kubernetes api, can update config new value in
replicas
field.
kubernetes has been designed auto-scaling handled external auto-scaler. discussed in responsibilities of replication controller in kubernetes docs.
Comments
Post a Comment