Get images from machine-os-images #220
Conversation
|
Skipping CI for Draft Pull Request. |
openshift-ci
bot
added
the
do-not-merge/work-in-progress
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: zaneb The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
openshift-ci
bot
added
the
approved
zaneb
force-pushed
the
machine-os-images
branch
4 times, most recently
from
3bb2cbb to
d18a036
Compare
|
/test all |
1 similar comment
|
/test all |
|
/test e2e-metal-ipi |
|
/test all |
zaneb
force-pushed
the
machine-os-images
branch
from
5e11830 to
ef784a2
Compare
|
/test e2e-metal-ipi-ovn-ipv6 |
|
@zaneb: The following tests failed, say
Full PR test history. Your PR dashboard. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here. |
|
/test all |
Pass images-bind-addr and images-publish-addr as env vars to the image customization controller.
Update CBO to ensure that the image customization deployment and service are up and deleted along with the metal3 deployment.
Pass the "-build-preprov-image" option to the BMO container to enable the Pre Provisioning Image feature. Also, stop setting the deployRamdiskUrl env var to stop BMO from starting its own internal controller.
The image publish address needs to be set on the command line, not in environment variables, so it was using the default (127.0.0.1:8084). The original idea behind using a NodePort Service was that it would be accessible to the BMCs. In practice it turns out we don't have a requirement for that yet - currently ironic is the only thing downloading images, and it caches them to serve up to the BMC. So for now downgrade this to a ClusterIP service (one of which is created by the NodePort Service anyway). We can always change it to NodePort if this becomes a requirement. The ClusterIP service can be referenced using the internal DNS name. To simplify URLs, expose the service on port 80 on the Pod's cluster IP.
The machine-os-images container supplies the latest ISO and PXE images for pre-provisioning from within the release payload. Use this instead of downloading preprovisioning images or using the RHEL IPA. Extract the images to the same location (and with the same names) in the image cache volume that we were using previously for IPA images. This change is transparent to the tasks using the images. Continue to run the second-level image cache service and download the Provisioning OS QCOW (if specified). Existing clusters installed prior to 4.10 will contain MachineSets that still rely on the QCOW for provisioning new Machines.
When booting an ISO, we set the ip=dhcp or ip=dhcp6 options inside the ISO image with the IP_OPTIONS env var in machine-os-images. When booting via PXE, we must set them on the PXE command line, by specifying the IRONIC_KERNEL_PARAMS env var in ironic.
zaneb
force-pushed
the
machine-os-images
branch
from
ef784a2 to
409d3c2
Compare
|
These changes were incorporated back into #208 |
|
@zaneb: PR needs rebase. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
openshift-ci
bot
added
the
needs-rebase
|
@zaneb: Closed this PR. DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
3 participants
No description provided.