argo-helm/charts/argo-workflows

Argo Workflows Chart

This is a community maintained chart. It is used to set up argo and its needed dependencies through one command. This is used in conjunction with helm.

If you want your deployment of this helm chart to most closely match the argo CLI, you should deploy it in the kube-system namespace.

Pre-Requisites

Custom resource definitions

Some users would prefer to install the CRDs outside of the chart. You can disable the CRD installation of this chart by using --set crds.install=false when installing the chart.

Helm cannot upgrade custom resource definitions in the <chart>/crds folder by design. Starting with 3.4.0 (chart version 0.19.0), the CRDs have been moved to <chart>/templates to address this design decision.

If you are using Argo Workflows chart version prior to 3.4.0 (chart version 0.19.0) or have elected to manage the Argo Workflows CRDs outside of the chart, please use kubectl to upgrade CRDs manually from templates/crds folder or via the manifests from the upstream project repo:

kubectl apply -k "https://github.com/argoproj/argo-workflows/manifests/base/crds/full?ref=<appVersion>"

# Eg. version v3.3.9
kubectl apply -k "https://github.com/argoproj/argo-workflows/manifests/base/crds/full?ref=v3.3.9"

ServiceAccount for Workflow Spec

In order for each Workflow run, you create ServiceAccount via values.yaml like below.

workflow:
  serviceAccount:
    create: true
    name: "argo-workflow"
  rbac:
    create: true
controller:
  workflowNamespaces:
    - default
    - foo
    - bar

Set ServiceAccount on Workflow.

apiVersion: argoproj.io/v1alpha1
kind: Workflow
metadata:
  generateName: hello-world-
spec:
  entrypoint: whalesay
  serviceAccountName: argo-workflow # Set ServiceAccount
  templates:
    - name: whalesay
      container:
        image: docker/whalesay
        command: [ cowsay ]
        args: [ "hello world" ]

Installing the Chart

To install the chart with the release name my-release:

$ helm repo add argo https://argoproj.github.io/argo-helm
"argo" has been added to your repositories

$ helm install my-release argo/argo-workflows
NAME: my-release
...

Changelog

For full list of changes, please check ArtifactHub changelog.

Usage Notes

Workflow controller

This chart defaults to setting the controller.instanceID.enabled to false now, which means the deployed controller will act upon any workflow deployed to the cluster. If you would like to limit the behavior and deploy multiple workflow controllers, please use the controller.instanceID.enabled attribute along with one of its configuration options to set the instanceID of the workflow controller to be properly scoped for your needs.

Workflow server authentication

By default, the chart requires some kind of authentication mechanism. This adopts the default behaviour from the Argo project itself. However, for local development purposes, or cases where your gateway authentication is covered by some other means, you can set the authentication mode for the Argo server by setting the server.extraArgs: [--auth-mode=server]. There are a few additional comments in the values.yaml file itself, including commented-out settings to disable authentication on the server UI itself using the same --auth-mode=server setting.

Values

The values.yaml contains items used to tweak a deployment of this chart. Fields to note:

• controller.instanceID.enabled: If set to true, the Argo Controller will ONLY monitor Workflow submissions with a --instanceid attribute
• controller.instanceID.useReleaseName: If set to true then chart set controller instance id to release name
• controller.instanceID.explicitID: Allows customization of an instance id for the workflow controller to monitor
• singleNamespace: When true, restricts the workflow controller to operate in just the single namespace (that one of the Helm release).
• controller.workflowNamespaces: This is a list of namespaces where the workflow controller will manage workflows. Only valid when singleNamespace is false.

General parameters

Key	Type	Default	Description
apiVersionOverrides.autoscaling	string	""	String to override apiVersion of autoscaling rendered by this helm chart
crds.annotations	object	{}	Annotations to be added to all CRDs
crds.install	bool	true	Install and upgrade CRDs
crds.keep	bool	true	Keep CRDs on chart uninstall
createAggregateRoles	bool	true	Create clusterroles that extend existing clusterroles to interact with argo-cd crds
emissary.images	list	[]	The command/args for each image on workflow, needed when the command is not specified and the emissary executor is used.
extraObjects	list	[]	Array of extra K8s manifests to deploy
fullnameOverride	string	nil	String to fully override "argo-workflows.fullname" template
images.pullPolicy	string	"Always"	imagePullPolicy to apply to all containers
images.pullSecrets	list	[]	Secrets with credentials to pull images from a private registry
images.tag	string	""	Common tag for Argo Workflows images. Defaults to .Chart.AppVersion.
kubeVersionOverride	string	""	Override the Kubernetes version, which is used to evaluate certain manifests
nameOverride	string	nil	String to partially override "argo-workflows.fullname" template
singleNamespace	bool	false	Restrict Argo to operate only in a single namespace (the namespace of the Helm release) by apply Roles and RoleBindings instead of the Cluster equivalents, and start workflow-controller with the --namespaced flag. Use it in clusters with strict access policy.

Workflow

Key	Type	Default	Description
workflow.namespace	string	nil	Deprecated; use controller.workflowNamespaces instead.
workflow.rbac.create	bool	true	Adds Role and RoleBinding for the above specified service account to be able to run workflows. A Role and Rolebinding pair is also created for each namespace in controller.workflowNamespaces (see below)
workflow.serviceAccount.annotations	object	{}	Annotations applied to created service account
workflow.serviceAccount.create	bool	false	Specifies whether a service account should be created
workflow.serviceAccount.labels	object	{}	Labels applied to created service account
workflow.serviceAccount.name	string	"argo-workflow"	Service account which is used to run workflows

Workflow Controller

Key	Type	Default	Description
controller.affinity	object	{}	Assign custom affinity rules
controller.clusterWorkflowTemplates.enabled	bool	true	Create a ClusterRole and CRB for the controller to access ClusterWorkflowTemplates.
controller.deploymentAnnotations	object	{}	deploymentAnnotations is an optional map of annotations to be applied to the controller Deployment
controller.extraArgs	list	[]	Extra arguments to be added to the controller
controller.extraContainers	list	[]	Extra containers to be added to the controller deployment
controller.extraEnv	list	[]	Extra environment variables to provide to the controller container
controller.extraInitContainers	list	[]	Enables init containers to be added to the controller deployment
controller.image.registry	string	"quay.io"	Registry to use for the controller
controller.image.repository	string	"argoproj/workflow-controller"	Registry to use for the controller
controller.image.tag	string	""	Image tag for the workflow controller. Defaults to .Values.images.tag.
controller.initialDelay	string	nil	Resolves ongoing, uncommon AWS EKS bug: https://github.com/argoproj/argo-workflows/pull/4224
controller.instanceID.enabled	bool	false	Configures the controller to filter workflow submissions to only those which have a matching instanceID attribute.
controller.instanceID.explicitID	string	""	Use a custom instanceID
controller.instanceID.useReleaseName	bool	false	Use ReleaseName as instanceID
controller.kubeConfig	object	{} (See values.yaml)	Configure when workflow controller runs in a different k8s cluster with the workflow workloads, or needs to communicate with the k8s apiserver using an out-of-cluster kubeconfig secret.
controller.links	list	[]	Configure Argo Server to show custom links
controller.livenessProbe	object	See values.yaml	Configure liveness probe for the controller
controller.loadBalancerSourceRanges	list	[]	Source ranges to allow access to service from. Only applies to service type LoadBalancer
controller.logging.format	string	"text"	Set the logging format (one of: text, json)
controller.logging.globallevel	string	"0"	Set the glog logging level
controller.logging.level	string	"info"	Set the logging level (one of: debug, info, warn, error)
controller.metricsConfig.enabled	bool	false	Enables prometheus metrics server
controller.metricsConfig.ignoreErrors	bool	false	Flag that instructs prometheus to ignore metric emission errors.
controller.metricsConfig.metricRelabelings	list	[]	ServiceMonitor metric relabel configs to apply to samples before ingestion
controller.metricsConfig.metricsTTL	string	""	How often custom metrics are cleared from memory
controller.metricsConfig.path	string	"/metrics"	Path is the path where metrics are emitted. Must start with a "/".
controller.metricsConfig.port	int	9090	Port is the port where metrics are emitted
controller.metricsConfig.portName	string	"metrics"	Container metrics port name
controller.metricsConfig.relabelings	list	[]	ServiceMonitor relabel configs to apply to samples before scraping
controller.metricsConfig.secure	bool	false	Flag that use a self-signed cert for TLS
controller.metricsConfig.servicePort	int	8080	Service metrics port
controller.metricsConfig.servicePortName	string	"metrics"	Service metrics port name
controller.metricsConfig.targetLabels	list	[]	ServiceMonitor will add labels from the service to the Prometheus metric
controller.name	string	"workflow-controller"	Workflow controller name string
controller.namespaceParallelism	string	nil	Limits the maximum number of incomplete workflows in a namespace
controller.navColor	string	""	Set ui navigation bar background color
controller.nodeEvents.enabled	bool	true	Enable to emit events on node completion.
controller.nodeSelector	object	{"kubernetes.io/os":"linux"}	Node selector
controller.parallelism	string	nil	parallelism dictates how many workflows can be running at the same time
controller.pdb.enabled	bool	false	Configure Pod Disruption Budget for the controller pods
controller.persistence	object	{}	enable persistence using postgres
controller.podAnnotations	object	{}	podAnnotations is an optional map of annotations to be applied to the controller Pods
controller.podGCDeleteDelayDuration	string	5s (Argo Workflows default)	The duration in seconds before the pods in the GC queue get deleted. A zero value indicates that the pods will be deleted immediately.
controller.podGCGracePeriodSeconds	string	30 seconds (Kubernetes default)	Specifies the duration in seconds before a terminating pod is forcefully killed. A zero value indicates that the pod will be forcefully terminated immediately.
controller.podLabels	object	{}	Optional labels to add to the controller pods
controller.podSecurityContext	object	{}	SecurityContext to set on the controller pods
controller.priorityClassName	string	""	Leverage a PriorityClass to ensure your pods survive resource shortages.
controller.rbac.accessAllSecrets	bool	false	Allows controller to get, list and watch all k8s secrets. Can only be used if secretWhitelist is empty.
controller.rbac.create	bool	true	Adds Role and RoleBinding for the controller.
controller.rbac.secretWhitelist	list	[]	Allows controller to get, list, and watch certain k8s secrets
controller.rbac.writeConfigMaps	bool	false	Allows controller to create and update ConfigMaps. Enables memoization feature
controller.replicas	int	1	The number of controller pods to run
controller.resourceRateLimit	object	{}	Globally limits the rate at which pods are created. This is intended to mitigate flooding of the Kubernetes API server by workflows with a large amount of parallel nodes.
controller.resources	object	{}	Resource limits and requests for the controller
controller.retentionPolicy	object	{}	Workflow retention by number of workflows
controller.securityContext	object	{"allowPrivilegeEscalation":false,"capabilities":{"drop":["ALL"]},"readOnlyRootFilesystem":true,"runAsNonRoot":true}	the controller container's securityContext
controller.serviceAccount.annotations	object	{}	Annotations applied to created service account
controller.serviceAccount.create	bool	true	Create a service account for the controller
controller.serviceAccount.labels	object	{}	Labels applied to created service account
controller.serviceAccount.name	string	""	Service account name
controller.serviceAnnotations	object	{}	Annotations to be applied to the controller Service
controller.serviceLabels	object	{}	Optional labels to add to the controller Service
controller.serviceMonitor.additionalLabels	object	{}	Prometheus ServiceMonitor labels
controller.serviceMonitor.enabled	bool	false	Enable a prometheus ServiceMonitor
controller.serviceMonitor.namespace	string	""	Prometheus ServiceMonitor namespace
controller.serviceType	string	"ClusterIP"	Service type of the controller Service
controller.telemetryConfig.enabled	bool	false	Enables prometheus telemetry server
controller.telemetryConfig.ignoreErrors	bool	false	Flag that instructs prometheus to ignore metric emission errors.
controller.telemetryConfig.metricsTTL	string	""	How often custom metrics are cleared from memory
controller.telemetryConfig.path	string	"/telemetry"	telemetry path
controller.telemetryConfig.port	int	8081	telemetry container port
controller.telemetryConfig.secure	bool	false	Flag that use a self-signed cert for TLS
controller.telemetryConfig.servicePort	int	8081	telemetry service port
controller.telemetryConfig.servicePortName	string	"telemetry"	telemetry service port name
controller.tolerations	list	[]	Tolerations for use with node taints
controller.topologySpreadConstraints	list	[]	Assign custom TopologySpreadConstraints rules to the workflow controller
controller.volumeMounts	list	[]	Additional volume mounts to the controller main container
controller.volumes	list	[]	Additional volumes to the controller pod
controller.workflowDefaults	object	{}	Default values that will apply to all Workflows from this controller, unless overridden on the Workflow-level. Only valid for 2.7+
controller.workflowNamespaces	list	["default"]	Specify all namespaces where this workflow controller instance will manage workflows. This controls where the service account and RBAC resources will be created. Only valid when singleNamespace is false.
controller.workflowRestrictions	object	{}	Restricts the Workflows that the controller will process. Only valid for 2.9+
controller.workflowWorkers	string	nil	Number of workflow workers

Workflow Main Container

Key	Type	Default	Description
mainContainer.env	list	[]	Adds environment variables for the Workflow main container
mainContainer.envFrom	list	[]	Adds reference environment variables for the Workflow main container
mainContainer.imagePullPolicy	string	""	imagePullPolicy to apply to Workflow main container. Defaults to .Values.images.pullPolicy.
mainContainer.resources	object	{}	Resource limits and requests for the Workflow main container
mainContainer.securityContext	object	{}	sets security context for the Workflow main container

Workflow Executor

Key	Type	Default	Description
executor.env	list	[]	Adds environment variables for the executor.
executor.image.pullPolicy	string	""	Image PullPolicy to use for the Workflow Executors. Defaults to .Values.images.pullPolicy.
executor.image.registry	string	"quay.io"	Registry to use for the Workflow Executors
executor.image.repository	string	"argoproj/argoexec"	Repository to use for the Workflow Executors
executor.image.tag	string	""	Image tag for the workflow executor. Defaults to .Values.images.tag.
executor.resources	object	{}	Resource limits and requests for the Workflow Executors
executor.securityContext	object	{}	sets security context for the executor container

Workflow Server

Key	Type	Default	Description
server.affinity	object	{}	Assign custom affinity rules
server.autoscaling.behavior	object	{}	Configures the scaling behavior of the target in both Up and Down directions. This is only available on HPA apiVersion autoscaling/v2beta2 and newer
server.autoscaling.enabled	bool	false	Enable Horizontal Pod Autoscaler ([HPA]) for the Argo Server
server.autoscaling.maxReplicas	int	5	Maximum number of replicas for the Argo Server [HPA]
server.autoscaling.minReplicas	int	1	Minimum number of replicas for the Argo Server [HPA]
server.autoscaling.targetCPUUtilizationPercentage	int	50	Average CPU utilization percentage for the Argo Server [HPA]
server.autoscaling.targetMemoryUtilizationPercentage	int	50	Average memory utilization percentage for the Argo Server [HPA]
server.baseHref	string	"/"	Value for base href in index.html. Used if the server is running behind reverse proxy under subpath different from /.
server.clusterWorkflowTemplates.enableEditing	bool	true	Give the server permissions to edit ClusterWorkflowTemplates.
server.clusterWorkflowTemplates.enabled	bool	true	Create a ClusterRole and CRB for the server to access ClusterWorkflowTemplates.
server.deploymentAnnotations	object	{}	optional map of annotations to be applied to the ui Deployment
server.enabled	bool	true	Deploy the Argo Server
server.extraArgs	list	[]	Extra arguments to provide to the Argo server binary, such as for disabling authentication.
server.extraContainers	list	[]	Extra containers to be added to the server deployment
server.extraEnv	list	[]	Extra environment variables to provide to the argo-server container
server.extraInitContainers	list	[]	Enables init containers to be added to the server deployment
server.image.registry	string	"quay.io"	Registry to use for the server
server.image.repository	string	"argoproj/argocli"	Repository to use for the server
server.image.tag	string	""	Image tag for the Argo Workflows server. Defaults to .Values.images.tag.
server.ingress.annotations	object	{}	Additional ingress annotations
server.ingress.enabled	bool	false	Enable an ingress resource
server.ingress.extraPaths	list	[]	Additional ingress paths
server.ingress.hosts	list	[]	List of ingress hosts
server.ingress.ingressClassName	string	""	Defines which ingress controller will implement the resource
server.ingress.labels	object	{}	Additional ingress labels
server.ingress.pathType	string	"Prefix"	Ingress path type. One of Exact, Prefix or ImplementationSpecific
server.ingress.paths	list	["/"]	List of ingress paths
server.ingress.tls	list	[]	Ingress TLS configuration
server.loadBalancerIP	string	""	Static IP address to assign to loadBalancer service type LoadBalancer
server.loadBalancerSourceRanges	list	[]	Source ranges to allow access to service from. Only applies to service type LoadBalancer
server.logging.format	string	"text"	Set the logging format (one of: text, json)
server.logging.globallevel	string	"0"	Set the glog logging level
server.logging.level	string	"info"	Set the logging level (one of: debug, info, warn, error)
server.name	string	"server"	Server name string
server.nodeSelector	object	{"kubernetes.io/os":"linux"}	Node selector
server.pdb.enabled	bool	false	Configure Pod Disruption Budget for the server pods
server.podAnnotations	object	{}	optional map of annotations to be applied to the ui Pods
server.podLabels	object	{}	Optional labels to add to the UI pods
server.podSecurityContext	object	{}	SecurityContext to set on the server pods
server.priorityClassName	string	""	Leverage a PriorityClass to ensure your pods survive resource shortages
server.rbac.create	bool	true	Adds Role and RoleBinding for the server.
server.replicas	int	1	The number of server pods to run
server.resources	object	{}	Resource limits and requests for the server
server.secure	bool	false	Run the argo server in "secure" mode. Configure this value instead of --secure in extraArgs.
server.securityContext	object	{"allowPrivilegeEscalation":false,"capabilities":{"drop":["ALL"]},"readOnlyRootFilesystem":false,"runAsNonRoot":true}	Servers container-level security context
server.serviceAccount.annotations	object	{}	Annotations applied to created service account
server.serviceAccount.create	bool	true	Create a service account for the server
server.serviceAccount.labels	object	{}	Labels applied to created service account
server.serviceAccount.name	string	""	Service account name
server.serviceAnnotations	object	{}	Annotations to be applied to the UI Service
server.serviceLabels	object	{}	Optional labels to add to the UI Service
server.serviceNodePort	string	nil	Service node port
server.servicePort	int	2746	Service port for server
server.servicePortName	string	""	Service port name
server.serviceType	string	"ClusterIP"	Service type for server pods
server.sso	object	{}	SSO configuration when SSO is specified as a server auth mode.
server.tolerations	list	[]	Tolerations for use with node taints
server.topologySpreadConstraints	list	[]	Assign custom TopologySpreadConstraints rules to the argo server
server.volumeMounts	list	[]	Additional volume mounts to the server main container.
server.volumes	list	[]	Additional volumes to the server pod.

Artifact Repository

Key	Type	Default	Description
artifactRepository.archiveLogs	bool	false	Archive the main container logs as an artifact
artifactRepository.azure	object	{} (See values.yaml)	Store artifact in Azure Blob Storage
artifactRepository.gcs	object	{} (See values.yaml)	Store artifact in a GCS object store
artifactRepository.s3	object	See values.yaml	Store artifact in a S3-compliant object store
customArtifactRepository	object	{}	The section of custom artifact repository. Utilize a custom artifact repository that is not one of the current base ones (s3, gcs, azure)
useStaticCredentials	bool	true	Use static credentials for S3 (eg. when not using AWS IRSA)

Breaking changes from the deprecated argo chart

1. the installCRD value has been removed. CRDs are now only installed from the conventional crds/ directory

2. the CRDs were updated to apiextensions.k8s.io/v1

3. the container image registry/project/tag format was changed to be more in line with the more common

   image:
     registry: quay.io
     repository: argoproj/argocli
     tag: v3.0.1
   

   this also makes it easier for automatic update tooling (eg. renovate bot) to detect and update images.

4. switched to quay.io as the default registry for all images

5. removed any included usage of Minio

6. aligned the configuration of serviceAccounts with the argo-cd chart, ie: what used to be server.createServiceAccount is now server.serviceAccount.create

7. moved the field previously known as telemetryServicePort inside the telemetryConfig as telemetryConfig.servicePort - same for metricsConfig