Use a pre-built node and master image #97
Comments
|
This issue was moved to kubernetes-sigs/cluster-api-provider-gcp#18 |
chuckha
pushed a commit
to chuckha/cluster-api
that referenced
this issue
Oct 2, 2019
Clean up verification tools
jayunit100
pushed a commit
to jayunit100/cluster-api
that referenced
this issue
Jan 31, 2020
…#124) Currently the configure part of the init script is also installing packages, which should not be the case. The install part of the script should be installing anything needed during the configure part. This split is further used when we want to use template that is pre-loaded with all the binaries. In that case, we only want to run the configure part which would bring up the services Resolves kubernetes-sigs#97 Resolves kubernetes-sigs#121 Change-Id: I720eaf54829dd3f756dbfedb7691c004c54fcd01
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
From @karan on January 24, 2018 19:18
Currently, spinning up nodes and masters can take ~minutes. Most of this time is due to apt-get installs. We can cache these by using a pre-built image, and then simply run
kubeadm join. This will make node startups faster.Similar for master.
https://github.com/kubernetes/kube-deploy/blob/master/cluster-api-gcp/cloud/google/templates.go#L159
Copied from original issue: kubernetes-retired/kube-deploy#514
The text was updated successfully, but these errors were encountered: