Update Machine Instances and Templates OCP+V#27
Update Machine Instances and Templates OCP+V#27tux1980 wants to merge 55 commits intostormshift:mainfrom
Conversation
Update HyperConverged.yaml stormshift#24 >> Correction of the structure - The correct indentation in the right place.
tux1980
left a comment
There was a problem hiding this comment.
Please Review and approve (Adding Templates and Instance Types + adapting the kustomize file.
Thanks
rbo
left a comment
There was a problem hiding this comment.
Please change the labels and base images. Details in comments.
There was a problem hiding this comment.
🤔 This VM should fail? IvyBridge-v2 is not RHEL 10 ready... Because Ivy do not support x86-64 ABI v3: https://qemu-project.gitlab.io/qemu/system/qemu-cpu-models.html
There was a problem hiding this comment.
Actually I deleted this yaml since it makes no sense days ago.... It shouldn't be there at all
| memory: | ||
| guest: "4Gi" | ||
| nodeSelector: | ||
| usage-broadwell: "yes" |
There was a problem hiding this comment.
Who created the label at the node?
% oc get nodes -l usage-broadwell
NAME STATUS ROLES AGE VERSION
storm5 Ready storage-node,virt-node,worker 237d v1.33.6
storm6 Ready storage-node,virt-node,worker 237d v1.33.6
It's not defined here: https://github.com/stormshift/clusters/blob/main/configuration/overlays/isar/cluster-scope/Node/storm6.yaml
Additionally, why not use the kubevirt one: cpu-model.node.kubevirt.io/Broadwell-v4
oc get nodes -l cpu-model.node.kubevirt.io/Broadwell-v4=true
NAME STATUS ROLES AGE VERSION
storm5 Ready storage-node,virt-node,worker 237d v1.33.6
storm6 Ready storage-node,virt-node,worker 237d v1.33.6
There was a problem hiding this comment.
Makes sense, using the kubevirt one.
cpu-model.node.kubevirt.io/Broadwell-v4=true
I# going to change that in the files, same for haswell
| memory: | ||
| guest: "4Gi" | ||
| nodeSelector: | ||
| usage-haswell: "yes" |
There was a problem hiding this comment.
Same as above. But different label: cpu-model.node.kubevirt.io/Haswell-v4
% oc get node -l cpu-model.node.kubevirt.io/Haswell-v4
NAME STATUS ROLES AGE VERSION
storm2 Ready virt-node,worker 237d v1.33.6
storm3 Ready storage-node,virt-node,worker 237d v1.33.6
storm5 Ready storage-node,virt-node,worker 237d v1.33.6
storm6 Ready storage-node,virt-node,worker 237d v1.33.6
There was a problem hiding this comment.
Why do we not use the DataSource:
% oc get datasource -n openshift-virtualization-os-images rhel10
NAME AGE
rhel10 196d
It's based on registry.redhat.io/rhel10/rhel-guest-image:latest as well?
You mentioned it in the call, but I forgot it.
There was a problem hiding this comment.
I wanted to have a new one since we had trouble with the metadata in the Original RHEL10 in connection with the SSP Operator. Furthermore, as you remember - we discussed to push all the customize stuff also in a new NS - "custom-os-images"
There was a problem hiding this comment.
Same as above:
Why do we not use the DataSource:
% oc get datasource -n openshift-virtualization-os-images rhel10
NAME AGE
rhel10 196d
There was a problem hiding this comment.
above - the ds - we need another set of Metadata - look in the Issue 312 please, there I was suppose to document.
There was a problem hiding this comment.
Same as above but for rhel9
% oc get datasource -n openshift-virtualization-os-images rhel9
NAME AGE
rhel9 2y66d
There was a problem hiding this comment.
above - the ds - we need another set of Metadata - look in the Issue 312 please, there I was suppose to document.
There was a problem hiding this comment.
Same as above but for rhel9
% oc get datasource -n openshift-virtualization-os-images rhel9
NAME AGE
rhel9 2y66d
There was a problem hiding this comment.
above - the ds - we need another set of Metadata - look in the Issue 312 please, there I was suppose to document.
There was a problem hiding this comment.
Not needed if we use the provided rhel9/10 images.
There was a problem hiding this comment.
We discussed to push all the customize stuff also in a new NS - "custom-os-images"
There was a problem hiding this comment.
But I forgot what is custom on that image?
There was a problem hiding this comment.
Maybe nothing, just wanted to make absolutely sure that we does not get any interference using images from the openshift-virtualization-os-images NS so that the SSP Operator maybe does not overwrite anything. So, we might change this again to the default image in the ds files:
source:
pvc:
name: rhel9-clean-base
namespace: custom-os-images
There was a problem hiding this comment.
But we should keep all other resources in the new NS "custom-os-images"- apart from the dv's. simply for clean separation
There was a problem hiding this comment.
I would prefer to use the datasource/rhel10 in openshift-virtualization-os-images
What is recommend from RH adding own images into openshift-virtualization-os-images or create own namespace?
There was a problem hiding this comment.
We might as well put it in openshift-virtualization-os-images. This was just our call - where we wanted to strictly distinguish between custom images and default images - even so it is the same, just to able to differ them visually.
| from: 'rhel9-[a-z0-9]{16}' | ||
| - name: DATA_SOURCE_NAME | ||
| description: Name of the DataSource to clone | ||
| value: rhel9 |
There was a problem hiding this comment.
BTW you do not use your own base images. You use the default one. 👍🏼 But then remove the copy of images.
There was a problem hiding this comment.
Yes, for the templates we didn#t need extra metadata like in the Datasources/Boot Volumes. Is this causing any harm to anyone?;)
Creating Instances and Templates via Git Ops for OCP Virtualization