ghcr.io/linuxsuren/argo-workflow-atomic-plugin:v0.0.1

此插件可减少不必要的 Argo 工作流。例如，短时间内对某个拉取请求有多个提交时，多数情况下只需保留最后一次工作流运行。此插件会停止所有具有相同参数且来自同一 WorkflowTemplate 的工作流。

安装

首先，启用 Argo Workflows 的插件功能：

kubectl patch deployment \
workflow-controller \
--namespace argo \
--type='json' \
-p='[{"op": "add", "path": "/spec/template/spec/containers/0/env/0", "value": {
"name": "ARGO_EXECUTOR_PLUGINS",
"value": "true",
}}]'

然后，将此插件安装为 ConfigMap：

cat <<EOF | kubectl apply -f -
---
apiVersion: v1
kind: ServiceAccount
metadata:
name: argo-atomic-plugin-executor-plugin
namespace: default
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
name: argo-plugin-addition-role
rules:
- apiGroups:
- argoproj.io
resources:
- workflowtasksets
- workflowtasksets/status
verbs:
- get
- watch
- patch
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: argo-plugin-atomic-addition-binding
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: argo-plugin-addition-role
subjects:
- kind: ServiceAccount
name: argo-atomic-plugin-executor-plugin
namespace: default
- kind: ServiceAccount
name: argo
namespace: argo
- kind: ServiceAccount
name: default
namespace: default
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: argo-plugin-atomic-binding
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: argo-server-cluster-role
subjects:
- kind: ServiceAccount
name: argo-atomic-plugin-executor-plugin
namespace: default
- kind: ServiceAccount
name: argo
namespace: argo
- kind: ServiceAccount
name: default
namespace: default
---
apiVersion: v1
data:
sidecar.automountServiceAccountToken: "true"
sidecar.container: |
image: ghcr.io/linuxsuren/argo-workflow-atomic-plugin:master
command:
- argo-wf-atomic
name: argo-atomic-plugin
ports:
- containerPort: 3002
resources:
limits:
cpu: 500m
memory: 128Mi
requests:
cpu: 250m
memory: 64Mi
securityContext:
allowPrivilegeEscalation: true
runAsNonRoot: true
runAsUser: 65534
kind: ConfigMap
metadata:
labels:
workflows.argoproj.io/configmap-type: ExecutorPlugin
name: argo-atomic-plugin
namespace: argo
EOF

试用

首先，创建 WorkflowTemplate：

cat <<EOF | kubectl apply -f -
apiVersion: argoproj.io/v1alpha1
kind: WorkflowTemplate
metadata:
name: plugin-atomic
namespace: default
spec:
entrypoint: main
templates:
- name: main
dag:
tasks:
- name: sleep
template: sleep
- name: atomic
template: atomic
- script:
image: ghcr.io/linuxsuren/hd:v0.0.70
command: [sh]
source: sleep 90
name: sleep
- name: atomic
plugin:
argo-atomic-plugin: {}
EOF

然后，通过 UI 或以下命令触发工作流：

cat <<EOF | kubectl create -f -
apiVersion: argoproj.io/v1alpha1
kind: Workflow
metadata:
generateName: plugin-atomic
namespace: default
labels:
workflows.argoproj.io/workflow-template: plugin-atomic
spec:
workflowTemplateRef:
name: plugin-atomic
EOF