If the provisioning of a kubernetes cluster fails but all the infrastructure components are correctly provisioned (i. e. virtual machines are up and running and joined to the cluster), it could be helpful to reinstall addon components such as CNI and CSI plugins.
Curious which release of Morpheus you are on. We added the ability to retry automation, post infrastructure deployment in either 6.0 or 6.0.1. I haven’t been able to cause a failure to ensure it also applies to cluster creation, however.
They are on 6.0.2
Hi @cbunge I am using version 6.0.2.
Can you tell me how to re-run automation post-provisioning?
Thanks in advance.
Under the history tab of either the cluster or the master/workers you should hopefully see the retry option on the task. But again, from a cluster perspective I would have to get one to fail to ensure that is applied at the same stage as you are seeing.
Hi Chris,
unfortunately there is no retry option.
I think this option should be always available.
Thanks
Stefano
Hi @mreed Any thoughts on this, is there a way to force the reinstallation of CNI and CSI in a Kubernetes cluster provisioned by Morpheus?
It appears currently we don’t support retry at that stage of a K8s deployment. This should be captured in ideas and the dev team can investigate if applicable.