Steps to install AWX using Kubernetes cluster.

Archives AWX Project
,
1

Hi
does anyone have step by step installation of AWX using Kubernetes cluster.

2

Straight from the README!

https://github.com/ansible/awx-operator

3

Github information mostly works if you can use the default images as is, requires bit reading and custom configuration to use or build your own image and run containers as non-root among other security considerations.

What issues are you having?

Thanks

4

we are getting these warnings when going through the configuration steps
[kubeadmin@xl8phxc360awxtst awx-operator-0.18.0]$ k get pods -A
NAMESPACE NAME READY STATUS RESTARTS AGE
awx awx-demo-postgres-0 0/1 Pending 0 5m8s
awx awx-operator-controller-manager-dfc6f67d7-4k7dp 2/2 Running 0 13m
kube-system coredns-64897985d-d5gg6 1/1 Running 1 (3d22h ago) 4d20h
kube-system coredns-64897985d-pqp2w 1/1 Running 1 (3d22h ago) 4d20h
kube-system etcd-xl8phxc360awxtst 1/1 Running 5 (3d22h ago) 4d20h
kube-system kube-apiserver-xl8phxc360awxtst 1/1 Running 5 (3d22h ago) 4d20h
kube-system kube-controller-manager-xl8phxc360awxtst 1/1 Running 1 (3d22h ago) 4d20h
kube-system kube-flannel-ds-z4vmf 1/1 Running 1 (3d22h ago) 4d20h
kube-system kube-proxy-2q497 1/1 Running 1 (3d22h ago) 4d20h
kube-system kube-scheduler-xl8phxc360awxtst 1/1 Running 5 (3d22h ago) 4d20h
[kubeadmin@xl8phxc360awxtst awx-operator-0.18.0]$ k describe pods awx-demo-postgres-0|grep -A5 Warning
Warning FailedScheduling 5m41s default-scheduler 0/1 nodes are available: 1 pod has unbound immediate PersistentVolumeClaims.
Warning FailedScheduling 3m33s (x1 over 4m33s) default-scheduler 0/1 nodes are available: 1 pod has unbound immediate PersistentVolumeClaims.
[kubeadmin@xl8phxc360awxtst awx-operator-0.18.0]$

5

Hello
here is a good project that helped me to install AWX on a k3s cluster.
https://github.com/kurokobo/awx-on-k3s

6

@karon they are using a K8s deployment, not K3s(minikube).

the link you posted is alright, however the deployment has been simplified greatly since, using kustomize.

as to OP, the error states you have no volumes available.

You need to create a PV/PVC for the postgres pod.

However, you’re likely going to run into the following, as I have. I am attempting to use localStorage/hostPath and it seems the postgres pod is failing to schedule.
I have the PV/PVC created for this, even the storage class is right. I’ve created 2 volumes, one for projects and one for postgres. projects is binding successfully, but postgres keeps staying in this state.

Been trying to sort this out for a week. This is a bare-metal cluster.

Events:
Type Reason Age From Message

Warning FailedScheduling 14s (x23 over 22m) default-scheduler 0/1 nodes are available: 1 node(s) didn’t find available persistent volumes to bind.