Upgrade rancher's kubernetes clusters to a supported version

https://kubernetes.io/releases/

1.22 is not supported since the 2022-10-28.
We should upgrade the clusters to 1.24 supported until 2023-11-09.
1.25 is not yet available in rancher.

Clusters to upgrade:

• archive-staging
• archive-production
• cluster-admin

added rancher state:wip labels

assigned to @vsellier

changed the description

List deployed applications to check if they are still compatible:

• metallb
• keda
• local-path-provisioner
• nginx-ingress (https://github.com/kubernetes/ingress-nginx/#supported-versions-table and https://artifacthub.io/packages/helm/ingress-nginx/ingress-nginx/4.2.5)
• swh
• reaper
• nfs-provisioner
• minio (?)
• cert-manager
• cert-manager-webhook-gandi
• argocd

keda looks ok until 1.25: https://github.com/kedacore/keda/issues/2462#issuecomment-1102046239

metallb 0.13.7 seems to be compatible with the 1.25. A upgrade of our version must be done as we are currently using the 0.13.5

nginx we are using the helm chart version 4.2.5 which deployed nginx v1.3.1 which is compatible until version 1.24. We will have to schedule an nginx upgrade before being able to upgrade to 1.25 but for 1.24 it's ok.

reaper simple application deployment should be ok for 1.24

minio found nothing

nfs-provisioner https://github.com/kubernetes-sigs/nfs-subdir-external-provisioner/tree/master/charts/nfs-subdir-external-provisioner#prerequisites

cert-manager https://artifacthub.io/packages/helm/cert-manager/cert-manager/1.10.0 (> 1.20+)

argocd no information provided

added #4689 (closed) as child task

List of deprecation messages in kibana (search for deprecated on the last 15 days in the systemlog indexes):

• minio: metallb.io v1beta1 AddressPool is deprecated, consider using IPAddressPool
• argocd / rancher: W1129 18:31:26.715352 1 warnings.go:70] policy/v1beta1 PodSecurityPolicy is deprecated in v1.21+, unavailable in v1.25+
• prometheus: W1129 18:31:04.816948 1 warnings.go:70] policy/v1beta1 PodDisruptionBudget is deprecated in v1.21+, unavailable in v1.25+; use policy/v1 PodDisruptionBudget
• prometheus: warnings.go:70] autoscaling/v2beta1 HorizontalPodAutoscaler is deprecated in v1.22+, unavailable in v1.25+; use autoscaling/v2beta2 HorizontalPodAutoscaler- rancher: W1129 18:32:13.939725 56 warnings.go:80] batch/v1beta1 CronJob is deprecated in v1.21+, unavailable in v1.25+; use batch/v1 CronJob
• rancher/canal: warnings.go:70] discovery.k8s.io/v1beta1 EndpointSlice is deprecated in v1.21+, unavailable in v1.25+; use discovery.k8s.io/v1 EndpointSlice

Everything is related to 1.25, not 1.24

mentioned in commit swh-sysadmin-provisioning@a036ef43

mentioned in merge request swh-sysadmin-provisioning!94 (merged)

mentioned in commit swh-sysadmin-provisioning@a905134d

marked the checklist item archive-staging as completed

marked the checklist item cluster-admin as completed

Admin cluster upgraded to 1.23:

kubectl get nodes                                                                                                                                                                                                         19:27:28
NAME                        STATUS   ROLES               AGE   VERSION
rancher-node-admin-mgmt1    Ready    controlplane,etcd   22d   v1.23.14
rancher-node-admin-node01   Ready    worker              22d   v1.23.14
rancher-node-admin-node02   Ready    worker              22d   v1.23.14
rancher-node-admin-node03   Ready    worker              22d   v1.23.14

archive-production cluster upgraded:

kubectl get nodes                                                                                                                                                                                                                19:58:00
NAME                               STATUS   ROLES               AGE   VERSION
rancher-node-production-mgmt1      Ready    controlplane,etcd   89d   v1.23.14
rancher-node-production-worker01   Ready    worker              89d   v1.23.14
rancher-node-production-worker02   Ready    worker              89d   v1.23.14
rancher-node-production-worker03   Ready    worker              89d   v1.23.14
rancher-node-production-worker04   Ready    worker              82d   v1.23.14
rancher-node-production-worker05   Ready    worker              50d   v1.23.14

marked the checklist item archive-production as completed

mentioned in commit swh-sysadmin-provisioning@b50d585f

mentioned in commit swh-sysadmin-provisioning@15f3ff73

mentioned in commit swh-sysadmin-provisioning@2804a8c0

removed state:wip label

closed

marked this issue as related to #4999 (closed)