Introduction
DCAE Platform in Dublin release supports new feature to deploy components via helm chart. This is enabled by integrating Helm Cloudify plugin into Cloudify Manager instance DCAE-Platform uses to deploy other required services. The cloudify plugin itself is under CCSDK project delivered part of Casablanca. For Dublin, this plugin has been integrated into DCAE ONAP deployment.
Dublin Scope
The helm plugin was intended to support deployment scenario of stand-alone application similar to capability offered under OOM. With this plugin integration, any charts packaged under ONAP OOM can be deployed through DCAE platform in ONAP. This provides an opportunity for operators to use a single orchestration through Cloudify for deploying both Helm and Toscal work flows if required.
As all DCAE MS currently are TOSCA work flow based, the helm plugin is not used for DCAE component deployment.
Artifacts
Repository Path: /org.onap.ccsdk.platform.plugins/plugins/helm-4.0.1-py27-none-linux_x86_64.wgn
Blueprint Template:
# ============LICENSE_START========================================== # =================================================================== # Copyright (c) 2019 AT&T # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. #============LICENSE_END============================================ tosca_definitions_version: cloudify_dsl_1_3 imports: - http://www.getcloudify.org/spec/cloudify/4.3.1/types.yaml - "https://nexus.onap.org/service/local/repositories/raw/content/org.onap.ccsdk.platform.plugins/type_files/helm/4.0.0/helm-type.yaml" inputs: tiller-server-ip: description: IP address of Kubernetes master node tiller-server-port: description: Nodeport of tiller server namespace: description: Target namespace to be installed under (requires to be new) chart-repo-url: default: https://nexus.onap.org/content/sites/oom-helm-staging chart-version : description: Chart version for identified component-name stable-repo-url: description: URL for stable repository type: string default: 'https://kubernetes-charts.storage.googleapis.com' config-url: default: '' config-format: default: 'yaml' component-name: description: onap component name node_templates: dcaecomponent: type: onap.nodes.component properties: tiller-server-ip: { get_input: tiller-server-ip } tiller-server-port: { get_input: tiller-server-port } component-name: { get_input: component-name } chart-repo-url: { get_input: chart-repo-url } chart-version: { get_input: chart-version } namespace: { get_input: namespace } stable-repo-url: { get_input: stable-repo-url} config-url: { get_input: config-url} config-format: { get_input: config-format} outputs: dcaecomponent_install_status: value: { get_attribute: [ dcaecomponent, install-status ] }
There is also option to override the defaults using values.yaml equivalent using following blueprint template
Pre-Configuration Steps
- Helm needs to be installed on the CM pod
kubectl exec -n onap <Cloudify Manager pod> /bin/bash wget http://storage.googleapis.com/kubernetes-helm/helm-v2.9.1-linux-amd64.tar.gz tar -zxvf helm-v2.9.1-linux-amd64.tar.gz sudo mv linux-amd64/helm /usr/local/bin/helm
2. Tiller service should be updated to expose a nodeport
kubectl edit svc -n kube-system tiller-deploy -o yaml # Assign an unused nodeport available in cluster
Installation
Both above templates are loaded into DCAE bootstrap pod for convenience
Note: Explanation of parameters are documented under CCSDK wiki page : Introduction of Helm Plugin
Future Enhancement
- Requires changing nodeport for tiller; since the deployment is done from bootstrap pod, clusterip/port must also be supportable
- Existing namespace cannot be used currently
- Deployment error not being logged