how to set up aws instances: instanceType and deviceMapping?

[hongkailiu]

Version

$ ./bin/openshift-install version
./bin/openshift-install v0.3.0-208-g54b49cdf0fb6d1150c7a9d4102e2b785d5a0344d
Terraform v0.11.8

Your version of Terraform is out of date! The latest version
is 0.11.10. You can update by downloading from www.terraform.io/downloads.html

Platform (aws|libvirt|openstack):

aws

What happened?

I would like to know how to set up the instance type and device mapping

$ bin/openshift-install create manifests
$ vi ./tectonic/99_openshift-cluster-api_master-machines.yaml
...
        instanceType: m5.xlarge

$ bin/openshift-install create cluster --log-level=debug

1. the modification on instanceType did not have impact. The created instances are still with the default t2.medium
2. need a way to set up device mapping

$ oc version
oc v4.0.0-alpha.0+a148deb-572
kubernetes v1.11.0+d4cacc0
features: Basic-Auth GSSAPI Kerberos SPNEGO

Server https://hongkliu-api.devcluster.openshift.com:6443
kubernetes v1.11.0+d4cacc0

$ oc explain machine
error: Couldn't find resource for "cluster.k8s.io/v1alpha1, Kind=Machine"

What you expected to happen?

Would like to see how to set those properties up for aws instances.

How to reproduce it (as minimally and precisely as possible)?

See above.

$ your-commands-here

Anything else we need to know?

This is more of questions instead of bug/issue report.
Would like to know how to do things with installer.

References

• enter text here.

[abhinavdahiya]

So you can edit install-config be following https://github.com/openshift/installer/blob/master/docs/user/overview.md#multiple-invocations

Definition for options supported by installer:
https://godoc.org/github.com/openshift/installer/pkg/types#InstallConfig

need a way to set up device mapping

I don't think installer supports that option.

[hongkailiu]

Thanks, @abhinavdahiya
It feel that I did the right steps for changing the instanceType (from https://github.com/openshift/installer/blob/master/docs/user/overview.md#multiple-invocations it is not clear which file I should edit for this purpose):

$ bin/openshift-install create manifests
$ vi ./tectonic/99_openshift-cluster-api_master-machines.yaml
...
        instanceType: m5.xlarge

$ bin/openshift-install create cluster --log-level=debug

However, the resulting cluster is still with t2.medium.
Let me know what I missed. Thanks.

[wking]

#792 should fix this.

[akrzos]

Do we know when this might be merged so we can change instancetype of the initially deployed machines?

[wking]

Worker handling might be fixed by #1076 if someone wants to give that a spin. #1088 is in flight with docs for adjusting these settings in the install-config.

[akrzos]

FWIW I can change the instancetype now:

Ex:

...
machines:
- name: master
  platform:
    aws:
      type: m5.large
  replicas: 3
- name: worker
  platform:
    aws:
      type: m5.large
  replicas: 3
...