Upgrade: 76-remove-armada-if-unused activation failed. Armada image not found

[David Barbosa Bastos]

Brief Description
-----------------
Execution of script 76 failed due to not correctly finding Armada images. This causes the upgrade-activate step to fail.

The armada image is not just airshipt/armada, you also need to remove starlingx/armada-image.

Severity
--------
Provide the severity of the defect.
Critical

Steps to Reproduce
------------------
From 22.12 Prestaging build, upgrade to master (23.09). This error is seen during controller 0 activation.

Expected Behavior
------------------
controller-0 should complete activation

Actual Behavior
----------------
controller-0 doesn't complete activation

Reproducibility
---------------
Reproducible

System Configuration
--------------------
SX and DX

Branch/Pull Time/Commit
-----------------------
SW_VERSION="23.09"
BUILD_TARGET="Unknown"
BUILD_TYPE="Informal"
BUILD_ID="n/a"

Last Pass
---------
https://review.opendev.org/c/starlingx/config/+/890651

Timestamp/Logs
--------------
[sysadmin@controller-0 ~(keystone_admin)]$ system registry-image-list
+----------------------------------------------------------+
| Image Name |
+----------------------------------------------------------+
| docker.io/fluxcd/helm-controller |
| docker.io/fluxcd/source-controller |
| docker.io/starlingx/armada-image |
| docker.io/starlingx/n3000-opae |
| docker.io/wind-river/cloud-platform-deployment-manager |
| docker.io/wind-river/dm-monitor |
| gcr.io/kubebuilder/kube-rbac-proxy |
| ghcr.io/helm/tiller |
| ghcr.io/k8snetworkplumbingwg/multus-cni |
| ghcr.io/k8snetworkplumbingwg/sriov-cni |
| ghcr.io/k8snetworkplumbingwg/sriov-network-device-plugin |
| k8s.gcr.io/coredns/coredns |
| k8s.gcr.io/defaultbackend-amd64 |
| k8s.gcr.io/etcd |
| k8s.gcr.io/ingress-nginx/controller |
| k8s.gcr.io/ingress-nginx/kube-webhook-certgen |
| k8s.gcr.io/kube-apiserver |
| k8s.gcr.io/kube-controller-manager |
| k8s.gcr.io/kube-proxy |
| k8s.gcr.io/kube-scheduler |
| k8s.gcr.io/pause |
| quay.io/calico/cni |
| quay.io/calico/kube-controllers |
| quay.io/calico/node |
| quay.io/jetstack/cert-manager-acmesolver |
| quay.io/jetstack/cert-manager-cainjector |
| quay.io/jetstack/cert-manager-controller |
| quay.io/jetstack/cert-manager-ctl |
| quay.io/jetstack/cert-manager-webhook |
| quay.io/k8scsi/snapshot-controller |
| quay.io/stackanetes/kubernetes-entrypoint |
| registry.k8s.io/defaultbackend-amd64 |
| registry.k8s.io/ingress-nginx/controller |
| registry.k8s.io/ingress-nginx/kube-webhook-certgen |
| registry.k8s.io/ingress-nginx/opentelemetry |
+----------------------------------------------------------+

Test Activity
-------------
Developer Testing

Workaround
----------
N/A

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/c/starlingx/config/+/904808
Committed: https://opendev.org/starlingx/config/commit/58c8f8ced6e461a8ee886c84dab5427377f3cfa1
Submitter: "Zuul (22348)"
Branch: master

commit 58c8f8ced6e461a8ee886c84dab5427377f3cfa1
Author: dbarbosa <email address hidden>
Date: Thu Jan 4 10:08:35 2024 -0300

    Armada image not found by script 76

    This change fixes the removal of Armada images from registry.
    Initially it was thought that it would only have the
    airshipt/armada image, but there is also starlingx/armada-image.
    To resolve this, the starlingx/armada-image key was added to
    the images to be deleted.

    Test Plan:
    PASS Build pkgs and build image
    PASS Upgrade SX stx-8 -> master
    PASS When the 76-remove-armada-if-unused.py script is run
    no error appears and at the end of activation no Armada
    image is found.

    Closes-Bug: 2048400

    Change-Id: I91ed6a2f8bd0478d845a68716ea6e19b978c47f9
    Signed-off-by: David Bastos <email address hidden>

[Ghada Khalil (gkhalil)]

Marking as Low priority as there is no official support for upgrades in stx; it's best effort only.