I can't Install Percona XtraDB Cluster on Minikube

MySQL & MariaDB
1

Dear All,

I can’t install Percona XtraDB Cluster on Minikube with the following options. I can’t connect to the db on any way.

I will very appreciate for any help!

Thanks: Bela

Instructions:

Windows: 21H2
Minikube:
minikube version: v1.27.0
commit: 4243041b7a72319b9be7842a7d34b6767bbdac2b

K8S:
$ k get all -ALL
NAMESPACE NAME READY STATUS RESTARTS AGE L
default pod/my-op-pxc-operator-84fd4c7b-fkksd 1/1 Running 2 (51m ago) 66m
default pod/percona-client 0/1 Error 0 46m
default pod/percona-xtradb-cluster-operator-85df56c9b5-klbt7 1/1 Running 0 50m
kube-system pod/coredns-565d847f94-n54jr 1/1 Running 2 (51m ago) 77m
kube-system pod/etcd-minikube 1/1 Running 1 (52m ago) 77m
kube-system pod/kube-apiserver-minikube 1/1 Running 1 (51m ago) 77m
kube-system pod/kube-controller-manager-minikube 1/1 Running 2 (51m ago) 77m
kube-system pod/kube-proxy-mdrk8 1/1 Running 2 77m
kube-system pod/kube-scheduler-minikube 1/1 Running 2 (51m ago) 77m
kube-system pod/storage-provisioner 1/1 Running 3 (51m ago) 77m
kubernetes-dashboard pod/dashboard-metrics-scraper-b74747df5-q6zjt 1/1 Running 2 (51m ago) 76m
kubernetes-dashboard pod/kubernetes-dashboard-54596f475f-zr74n 1/1 Running 2 (51m ago) 76m

NAMESPACE NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE L
default service/kubernetes ClusterIP 10.96.0.1 443/TCP 77m
default service/percona-xtradb-cluster-operator ClusterIP 10.108.45.116 443/TCP 66m
kube-system service/kube-dns ClusterIP 10.96.0.10 53/UDP,53/TCP,9153/TCP 77m
kubernetes-dashboard service/dashboard-metrics-scraper ClusterIP 10.108.6.193 8000/TCP 76m
kubernetes-dashboard service/kubernetes-dashboard ClusterIP 10.101.49.179 80/TCP 76m

NAMESPACE NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE L
kube-system daemonset.apps/kube-proxy 1 1 1 1 1 kubernetes.io/os=linux 77m

NAMESPACE NAME READY UP-TO-DATE AVAILABLE AGE L
default deployment.apps/my-op-pxc-operator 1/1 1 1 66m
default deployment.apps/percona-xtradb-cluster-operator 1/1 1 1 50m
kube-system deployment.apps/coredns 1/1 1 1 77m
kubernetes-dashboard deployment.apps/dashboard-metrics-scraper 1/1 1 1 76m
kubernetes-dashboard deployment.apps/kubernetes-dashboard 1/1 1 1 76m

NAMESPACE NAME DESIRED CURRENT READY AGE L
default replicaset.apps/my-op-pxc-operator-84fd4c7b 1 1 1 66m
default replicaset.apps/percona-xtradb-cluster-operator-85df56c9b5 1 1 1 50m
kube-system replicaset.apps/coredns-565d847f94 1 1 1 77m
kubernetes-dashboard replicaset.apps/dashboard-metrics-scraper-b74747df5 1 1 1 76m
kubernetes-dashboard replicaset.apps/kubernetes-dashboard-54596f475f 1 1 1 76m

2

This:
$ kubectl apply -f https://raw.githubusercontent.com/percona/percona-xtradb-cluster-operator/v1.11.0/deploy/cr-minimal.yaml

…doesn’t deploy HAProxy node.

1 Like
3

The logs:

Operator:
DB: DBLog.txt - Google Drive

1 Like
4

Hello @Beci_Roboz ,

you run minikube 1.27, which is based on k8s v1.25.

In 1.25 PodDisruptionBudget beta API was removed (deprecated in 1.21).

PXC Operator 1.11 was released before k8s 1.25 and we have not changed the API for PDBs. As a result you see this error in the logs:

{"level":"error","ts":1664507151.3384674,"logger":"controller.perconaxtradbcluster-controller","msg":"Reconciler error","name":"minimal-cluster","namespace":"default","error":"PodDisruptionBudget for minimal-cluster-pxc: reconcile pdb: get object: no matches for kind \"P
odDisruptionBudget\" in version \"policy/v1beta1\"","errorVerbose":"no matches for kind \"PodDisruptionBudget\" in version \"policy/v1beta1

We state which versions of k8s were tested with the Operator in release notes.

As a workaround please use k8s < 1.25 for now.
We will release the new version of the Operator this quarter (Q4’2023). This ticket covers the PDBs: [K8SPXC-983] Update PodDisruptionBudget to policy/v1 - Percona JIRA

Please let me know if it helps.

2 Likes
5

Dear @Sergey_Pronin,

OK I will keep an eye on that.

Thank you so much!

Bela

1 Like