This is a first stab at vmhost-x86-01.rdu3.
It's already been installed with our kickstart.
The network config is just my best guess/first stab.
It's 2 nic's as bond devices, with a bridge on top.
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
This is a follow up of the migration to Deployment. Based on learnings
from a successful coreos-ostree-pruner migration, this fixes the
`no image` and `unknown field \"spec.strategy.resources\` errors
presented on batcave.
This is a follow up of the migration to Deployment. Based on learnings
from a succesfull coreos-cincinnati modification, this fixes the
`no image` error displayed currently on batcave.
As the previous change took no effect and the error while executing
`sudo rbac-playbook -l os_control_stg openshift-apps/coreos-cincinnati.yml`
continues to return an error as below, let's try another aproach.
```
"stderr": "The Deployment \"coreos-cincinnati\" is invalid: \n*
spec.template.spec.containers[0].image: Required value
```
Here adding a specific F42 image
During recent migration to deployment, the oc command:
`sudo rbac-playbook -l os_control_stg openshift-apps/coreos-cincinnati.yml`
returns an error:
```
"stderr": "The Deployment \"coreos-cincinnati\" is invalid: \n*
spec.template.spec.containers[0].image: Required value
```
This PR aims to fix the above.
When running
`sudo rbac-playbook -l os_control_stg openshift-apps/coreos-koji-tagger.yml`
An error is present as a result of not complete migration to deployment:
strict decoding error: unknown field \"spec.strategy.resources\"
This aims to fix the above.
When running
`sudo rbac-playbook -l os_control_stg openshift-apps/coreos-cincinnati.yml`
I came across the below errors due to recent migration to deployment and j2.
unknown field :\"spec.strategy
.activeDeadlineSeconds /.recreateParams /.resources /.rollingParams /.name
This should fix all of the above.
In order to correctly sync each architecture's repos, we have to
override the basearch while running reposync. Otherwise we are just
syncing the the host architecture (x86_64) multiple times under
misleading directory names.
Signed-off-by: Carl George <carlwgeorge@gmail.com>
Zstd repodata support was introduced in DNF quite some time ago, and
is quite a bit faster and less resource intensive.
Additionally, there is no point in producing sqlmd or DeltaRPMs when
support for both has been removed from RHEL for a fair bit of time.
Regardless, we should leave things mostly as-is for EPEL 9, and thus
we retain older xz compression settings for that.
Signed-off-by: Neal Gompa <ngompa@fedoraproject.org>
This should setup the rdu3 vlans with initial dns, which is just a copy
of our iad2 networks with 10.3 changing to 10.16.
Once rdu3 machines have network we should be able to use these to let
rdu3 hosts query the regular dns servers for things.
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
It turns out the ostree volumes are special in that they are NFS shares
mapped to the NetApp Volumes for OSTree repo storage. They have
specific names and require no storage class to be set. This reverts the
changes made yesterday.