Closed plegg-rh closed 1 year ago
plegg-rh
commented
1 year ago helm version
version.BuildInfo{Version:"v3.7.1+7.el8", GitCommit:"8f33223fe17957f11ba7a88b016bc860f034c4e6", GitTreeState:"clean", GoVersion:"go1.16.7"}
PyYAML version
Name: PyYAML Version: 6.0 Summary: YAML parser and emitter for Python Home-page: https://pyyaml.org/ Author: Kirill Simonov Author-email: xi@resolvent.net License: MIT Location: /usr/local/lib64/python3.6/site-packages Requires:
gravesm
commented
1 year ago @plegg-rh I'm unable to reproduce this. Could you provide the role that's being included?
plegg-rh
commented
1 year ago @gravesm , thanks for the quick reply
helm > tasks > main.yaml
---
- name: "Ensure repo ({{ helm.repo_name }}) is added to Helm for {{ helm.release_name }}"
kubernetes.core.helm_repository:
name: "{{ helm.repo_name }}"
repo_url: "{{ helm.chart_repo_url }}"
- name: "Configure release {{ helm.release_name }} on {{ ocp_cluster.env }} Cluster"
kubernetes.core.helm:
api_key: "{{ login_info.openshift_auth.api_key }}"
host: "https://api.{{ ocp_cluster.env }}.{{ install_config.base_domain }}:6443"
validate_certs: false
release_state: present
chart_ref: "{{ helm.chart_ref }}"
name: "{{ helm.release_name }}"
release_namespace: "{{ helm.release_namespace }}"
chart_repo_url: "{{ helm.chart_repo_url }}"
chart_version: "{{ helm.chart_version }}"
create_namespace: true
update_repo_cache: true
values_files:
- "{{ helm.chart_values }}"
helm > default > main.yaml
---
helm:
release_name: cert-manager
release_namespace: cert-manager
chart_repo_url: https://charts.jetstack.io/
chart_version: 1.8.2
create_namespace: true
update_repo_cache: true
chart_values: "{{ playbook_dir }}/helm-values/cert-manager-values.yaml"
repo_name: jetstack
playbook:
---
- name: Get host info for base plays
hosts: localhost
gather_facts: false
vars_files:
- charts.yaml
tasks:
- name: Login to cluster
ansible.builtin.import_role:
name: login_to_cluster
- name: Deploy Helm Charts
ansible.builtin.include_role:
name: helm
vars:
helm:
release_name: "{{ item.release_name }}"
release_namespace: "{{ item.release_namespace }}"
chart_repo_url: "{{ item.chart_repo_url }}"
chart_version: "{{ item.chart_version }}"
chart_values: "{{ item.chart_values }}"
repo_name: "{{ item.repo_name }}"
chart_ref: "{{ item.chart_ref }}"
loop: "{{ charts }}"I've tried again in a venv with the following packages, and still getting the same result.
Package Version
------------------- --------
attrs 22.2.0
cachetools 4.2.4
certifi 2023.5.7
charset-normalizer 2.0.12
google-auth 2.20.0
idna 3.4
importlib-metadata 4.8.3
jsonschema 3.2.0
kubernetes 26.1.0
kubernetes-validate 1.26.0
oauthlib 3.2.2
openshift 0.13.1
pip 21.3.1
pyasn1 0.5.0
pyasn1-modules 0.3.0
pyrsistent 0.18.0
python-dateutil 2.8.2
python-string-utils 1.0.0
PyYAML 6.0
requests 2.27.1
requests-oauthlib 1.3.1
rsa 4.9
setuptools 39.2.0
six 1.16.0
typing_extensions 4.1.1
urllib3 1.26.16
websocket-client 1.3.1
zipp 3.6.0I tried again to reproduce this using python 3.9, ansible 2.13 and kubernetes.core 2.4. With the role you provided I'm able to successfully deploy multiple helm charts into openshift. One thing I noticed from the exception you posted is that the module is being executed with python 3.6, though you list python 3.9 in your ansible version output. I also tried reproducing with python 3.6 but it still works for me. It might be worth setting ansible_python_interpreter just to be absolutely certain which python interpreter and environment is being used.
plegg-rh
commented
1 year ago I was able to succeed by using python3.6 interpreter on this machine.
I was also able to get the playbook to work in a ubi8 container with python3.9.
TL;DR it looks like this is a machine specific issue.
GuilhermeCamposo
commented
2 months ago I'm facing the same problem.
ansible [core 2.17.3] python version = 3.12.4 (main, Jun 7 2024, 00:00:00) [GCC 14.1.1 20240607 (Red Hat 14.1.1-5)] (/usr/bin/python3) jinja version = 3.1.4
Collection Version
kubernetes.core 5.0.0
name: Deploy redhat-trusted-profile-analyzer chart kubernetes.core.helm: api_key: '{{ token }}' host: '{{ server }}' validate_certs: false name: redhat-trusted-profile-analyzer chart_ref: openshift-helm-charts/redhat-trusted-profile-analyzer chart_version: '0.1.1' release_namespace: '{{ tpa_project }}' update_repo_cache: false values: "{{ lookup('template', 'values-rhtpa.yaml.j2') | from_yaml }}"
Any idea how I can debug this further?
SUMMARY
When using kubernetes.core.helm to release a chart on the cluster, the task fails with
'AnsibleModule' object has no attribute 'env_update'
ISSUE TYPE
COMPONENT NAME
kubernetes.core.helm
ANSIBLE VERSION
COLLECTION VERSION
CONFIGURATION
OS / ENVIRONMENT
RHEL 8.7
STEPS TO REPRODUCE
Call the kubernetes.core.helm module with the following params:
invocation: api_key:
chart_ref: jetstack/cert-manager
chart_version: v1.8.2
create_namespace: true
host:
module_args:
api_key:
chart_ref: jetstack/cert-manager
chart_version: v1.8.2
create_namespace: true
host:
release_name: cert-manager
release_namespace: cert-manager
release_state: present
validate_certs: false
values_files: ~/oke-auto-deploy/ipi-cluster-install/ansible/helm-values/cert-manager-values.yaml
release_name: cert-manager
release_namespace: cert-manager
release_state: present
validate_certs: false
values_files: ~/oke-auto-deploy/ipi-cluster-install/ansible/helm-values/cert-manager-values.yaml
EXPECTED RESULTS
helm chart jetstack/cert-manager-v1.8.2 would be deployed to host with the values file.
ACTUAL RESULTS
Somewhere an attribute is being defined, by doesn't exist in helm.py.