Failed HelmRelease is not redeployed

Context

I faced a failure with the following error during atmosphere converge(without any tagging, just ran the full playbook) poetry run molecule converge -s default.

TASK [vexxhost.atmosphere.neutron : Deploy Helm chart] *************************
fatal: [ctl1]: FAILED! => {"changed": false, "command": "/usr/bin/helm upgrade -i --reset-values --create-namespace -f=/tmp/tmpxovxqef8.yml neutron /usr/local/src/neutron", "msg": "Failure when executing Helm command. Exited 1.\nstdout: Release \"neutron\" does not exist. Installing it now.\n\nstderr: Error: secrets \"neutron-etc\" already exists\n", "stderr": "Error: secrets \"neutron-etc\" already exists\n", "stderr_lines": ["Error: secrets \"neutron-etc\" already exists"], "stdout": "Release \"neutron\" does not exist. Installing it now.\n", "stdout_lines": ["Release \"neutron\" does not exist. Installing it now."]}

I jumped into ctl1 node and checked the HelmRelease status and it was failed.

root@ctl1:/home/ubuntu# helm status neutron
NAME: neutron
LAST DEPLOYED: Wed Jul  5 14:47:41 2023
NAMESPACE: openstack
STATUS: failed
REVISION: 1

I tried to redeploy neutron poetry run molecule converge -s default -- -- --tags neutron. This time the chart deployment task passed without any error but HelmRelease was neither updated or replaced. Therefore finally failed at Wait until network service ready task.

TASK [vexxhost.atmosphere.neutron : Deploy Helm chart] *************************
ok: [ctl1]

TASK [Create Ingress] **********************************************************

TASK [vexxhost.atmosphere.openstack_helm_ingress : Create certificate] *********
skipping: [ctl1]

TASK [vexxhost.atmosphere.openstack_helm_ingress : Set fact with wildcard certificate] ***
skipping: [ctl1]

TASK [vexxhost.atmosphere.openstack_helm_ingress : Add ClusterIssuer annotations] ***
ok: [ctl1]

TASK [vexxhost.atmosphere.openstack_helm_ingress : Create Ingress network] *****
changed: [ctl1]

TASK [vexxhost.atmosphere.neutron : Wait until network service ready] **********
# stuck here and timed out

Expected result

Neutron HelmRelease should be redeployed.

Workaround

Need to remove failed HelmRelease manually and rerun the playbook.

Solution?

Helm chart deployment task is performed by kubernetes.core.helm module. Need to get the status of HelmRelease if exists already and if it is under failed status, do redeploy or sth.

vexxhost / atmosphere