On Tue, 19 Nov 2019 at 14:34, Nathanaël Blanchet <blanchet@abes.fr> wrote:
Le 19/11/2019 à 08:55, Roy Golan a écrit :
oc get -o json clusterversion 

This is the output of the previous failed deployment, I'll give a try to a newer one when I'll have a minute to test

(do I need to use the terraform-workers tag instead of latest?)

docker pull quay.io/rgolangh/openshift-installer:terraform-workers

[root@openshift-installer openshift-origin-client-tools-v3.11.0-0cbc58b-linux-64bit]# ./oc get -o json clusterversion
{
    "apiVersion": "v1",
    "items": [
        {
            "apiVersion": "config.openshift.io/v1",
            "kind": "ClusterVersion",
            "metadata": {
                "creationTimestamp": "2019-11-07T12:23:06Z",
                "generation": 1,
                "name": "version",
                "namespace": "",
                "resourceVersion": "3770202",
                "selfLink": "/apis/config.openshift.io/v1/clusterversions/version",
                "uid": "77600bba-6e71-4b35-a60b-d8ee6e0f545c"
            },
            "spec": {
                "channel": "stable-4.3",
                "clusterID": "6f87b719-e563-4c0b-ab5a-1144172bc983",
                "upstream": "https://api.openshift.com/api/upgrades_info/v1/graph"
            },
            "status": {
                "availableUpdates": null,
                "conditions": [
                    {
                        "lastTransitionTime": "2019-11-07T12:23:12Z",
                        "status": "False",
                        "type": "Available"
                    },
                    {script
                        "lastTransitionTime": "2019-11-07T12:56:15Z",
                        "message": "Cluster operator image-registry is still updating",
                        "reason": "ClusterOperatorNotAvailable",
                        "status": "True",
                        "type": "Failing"
                    },
                    {
                        "lastTransitionTime": "2019-11-07T12:23:12Z",
                        "message": "Unable to apply 4.3.0-0.okd-2019-10-29-180250: the cluster operator image-registry has not yet successfully rolled out",
                        "reason": "ClusterOperatorNotAvailable",
                        "status": "True",
                        "type": "Progressing"
                    },
                    {
                        "lastTransitionTime": "2019-11-07T12:23:12Z",
                        "message": "Unable to retrieve available updates: currently installed version 4.3.0-0.okd-2019-10-29-180250 not found in the \"stable-4.3\" channel",
                        "reason": "RemoteFailed",
                        "status": "False",
                        "type": "RetrievedUpdates"
                    }
                ],
                "desired": {
                    "force": false,
                    "image": "registry.svc.ci.openshift.org/origin/release@sha256:68286e07f7d68ebc8a067389aabf38dee9f9b810c5520d6ee4593c38eb48ddc9",
                    "version": "4.3.0-0.okd-2019-10-29-180250"

Indeed this version is not the latest and is missing the aforementioned fix for the registry.

                },
                "history": [
                    {
                        "completionTime": null,
                        "image": "registry.svc.ci.openshift.org/origin/release@sha256:68286e07f7d68ebc8a067389aabf38dee9f9b810c5520d6ee4593c38eb48ddc9",
                        "startedTime": "2019-11-07T12:23:12Z",
                        "state": "Partial",
                        "verified": false,
                        "version": "4.3.0-0.okd-2019-10-29-180250"
                    }
                ],
                "observedGeneration": 1,
                "versionHash": "-3onP9QpPTg="
            }
        }
    ],
    "kind": "List",
    "metadata": {
        "resourceVersion": "",
        "selfLink": ""
    }

}


Can you answer to these few questions please?

  • The latest stable OKD version is 4.2.4. Is it possible to chose the version of okd when deploying (seems to use 4.3) or does the installer always download the latest OKD?
 
  • Can we use FCOS instead of RHCOS?

  • About the pull secret, do we absolutely need a redhat login to get this file to deploy an upstream OKD cluster and not downstream openshift?

To answer the 3 of those, this specific build is not really OKD, and will use 4.3 and Red Hat artifact and must use RHCOs, hence the pull secret thing.
I frankly don't know when OKD 4.3 is going to be released, I guess it will be on top FCOS.
I'll update the list once we have the oVirt installer for OKD ready for testing (on FCOS)



-- 
Nathanaël Blanchet

Supervision réseau
Pôle Infrastrutures Informatiques
227 avenue Professeur-Jean-Louis-Viala
34193 MONTPELLIER CEDEX 5 	
Tél. 33 (0)4 67 54 84 55
Fax  33 (0)4 67 54 84 14
blanchet@abes.fr