clearml-helm-charts/charts/clearml

ClearML Ecosystem for Kubernetes

MLOps platform

Homepage: https://clear.ml

Maintainers

Name	Email	Url
valeriano-manassero		https://github.com/valeriano-manassero

Introduction

The clearml-server is the backend service infrastructure for ClearML. It allows multiple users to collaborate and manage their experiments.

clearml-server contains the following components:

• The ClearML Web-App, a single-page UI for experiment management and browsing
• RESTful API for:
  • Documenting and logging experiment information, statistics and results
  • Querying experiments history, logs and results
• Locally-hosted file server for storing images and models making them easily accessible using the Web-App

Local environment

For development/evaluation it's possible to use kind. After installation, following commands will create a complete ClearML insatllation:

cat <<EOF | kind create cluster --config=-                                                                  ─╯
kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
  extraPortMappings:
  # API server's default nodePort is 30008. If you customize it in helm values by
  # `apiserver.service.nodePort`, `containerPort` should match it
  - containerPort: 30008
    hostPort: 30008
    listenAddress: "127.0.0.1"
    protocol: TCP
  # Web server's default nodePort is 30080. If you customize it in helm values by
  # `webserver.service.nodePort`, `containerPort` should match it
  - containerPort: 30080
    hostPort: 30080
    listenAddress: "127.0.0.1"
    protocol: TCP
  # File server's default nodePort is 30081. If you customize it in helm values by
  # `fileserver.service.nodePort`, `containerPort` should match it
  - containerPort: 30081
    hostPort: 30081
    listenAddress: "127.0.0.1"
    protocol: TCP
  extraMounts:
  - hostPath: /tmp/clearml-kind/
    containerPath: /var/local-path-provisioner
EOF

helm install clearml allegroai/clearml

After deployment, the services will be exposed on localhost on the following ports:

• API server on 30008
• Web server on 30080
• File server on 30081

Data persisted in every Kubernetes volume by ClearML will be accessible in /tmp/clearml-kind folder on the host.

Production cluster environment

In a production environment it's suggested to install an ingress controller and verify that is working correctly. During ClearML deployment enable ingress section of chart values. This will create 3 ingress rules:

• app.<your domain name>
• files.<your domain name>
• api.<your domain name>

(for example, app.clearml.mydomainname.com, files.clearml.mydomainname.com and api.clearml.mydomainname.com)

Just pointing the domain records to the IP where ingress controller is responding will complete the deployment process.

Upgrades/ Values upgrades

Updating to latest version of this chart can be done in two steps:

helm repo update
helm upgrade clearml allegroai/clearml

Changing values on existing installation can be done with:

helm upgrade clearml allegroai/clearml --version <CURRENT CHART VERSION> -f custom_values.yaml

Please note: updating values only should always be done setting explicit chart version to avoid a possible chart update. Keeping separate updates procedures between version and values can be a good practice to seprate potential concerns.

Additional Configuration for ClearML Server

You can also configure the clearml-server for:

• fixed users (users with credentials)
• non-responsive experiment watchdog settings

For detailed instructions, see the Optional Configuration section in the clearml-server repository README file.

Source Code

• https://github.com/allegroai/clearml-helm-charts
• https://github.com/allegroai/clearml

Requirements

Kubernetes: >= 1.21.0-0 < 1.26.0-0

Repository	Name	Version
file://../../dependency_charts/elasticsearch	elasticsearch	7.16.2
file://../../dependency_charts/mongodb	mongodb	10.3.4
file://../../dependency_charts/redis	redis	10.9.0

Values

Key	Type	Default	Description
apiserver.affinity	object	{}
apiserver.authCookiesMaxAge	int	864000	Amount of seconds the authorization cookie will last in user browser
apiserver.configDir	string	"/opt/clearml/config"
apiserver.configuration	object	{"additionalConfigs":{},"configRefName":"","secretRefName":""}	additional configurations that can be used by api server; check examples in values.yaml file
apiserver.extraEnvs	list	[]
apiserver.image.pullPolicy	string	"IfNotPresent"
apiserver.image.repository	string	"allegroai/clearml"
apiserver.image.tag	string	"1.7.0"
apiserver.livenessDelay	int	60
apiserver.nodeSelector	object	{}
apiserver.podAnnotations	object	{}
apiserver.prepopulateArtifactsPath	string	"/mnt/fileserver"
apiserver.prepopulateEnabled	string	"true"
apiserver.prepopulateZipFiles	string	"/opt/clearml/db-pre-populate"
apiserver.readinessDelay	int	60
apiserver.replicaCount	int	1
apiserver.resources	object	{}
apiserver.service.nodePort	int	30008	If service.type set to NodePort, this will be set to service's nodePort field. If service.type is set to others, this field will be ignored
apiserver.service.port	int	8008
apiserver.service.type	string	"NodePort"	This will set to service's spec.type field
apiserver.tolerations	list	[]
clearml	object	{"defaultCompany":"d1bd92a3b039400cbafc60a7a5b1e52b"}	ClearMl generic configurations
elasticsearch.clusterHealthCheckParams	string	"wait_for_status=yellow&timeout=1s"
elasticsearch.clusterName	string	"clearml-elastic"
elasticsearch.enabled	bool	true
elasticsearch.esConfig."elasticsearch.yml"	string	"xpack.security.enabled: false\n"
elasticsearch.esJavaOpts	string	"-Xmx2g -Xms2g"
elasticsearch.extraEnvs[0].name	string	"bootstrap.memory_lock"
elasticsearch.extraEnvs[0].value	string	"false"
elasticsearch.extraEnvs[1].name	string	"cluster.routing.allocation.node_initial_primaries_recoveries"
elasticsearch.extraEnvs[1].value	string	"500"
elasticsearch.extraEnvs[2].name	string	"cluster.routing.allocation.disk.watermark.low"
elasticsearch.extraEnvs[2].value	string	"500mb"
elasticsearch.extraEnvs[3].name	string	"cluster.routing.allocation.disk.watermark.high"
elasticsearch.extraEnvs[3].value	string	"500mb"
elasticsearch.extraEnvs[4].name	string	"cluster.routing.allocation.disk.watermark.flood_stage"
elasticsearch.extraEnvs[4].value	string	"500mb"
elasticsearch.extraEnvs[5].name	string	"http.compression_level"
elasticsearch.extraEnvs[5].value	string	"7"
elasticsearch.extraEnvs[6].name	string	"reindex.remote.whitelist"
elasticsearch.extraEnvs[6].value	string	"*.*"
elasticsearch.extraEnvs[7].name	string	"xpack.monitoring.enabled"
elasticsearch.extraEnvs[7].value	string	"false"
elasticsearch.extraEnvs[8].name	string	"xpack.security.enabled"
elasticsearch.extraEnvs[8].value	string	"false"
elasticsearch.httpPort	int	9200
elasticsearch.minimumMasterNodes	int	1
elasticsearch.persistence.enabled	bool	true
elasticsearch.replicas	int	1
elasticsearch.resources.limits.memory	string	"4Gi"
elasticsearch.resources.requests.memory	string	"4Gi"
elasticsearch.roles.data	string	"true"
elasticsearch.roles.ingest	string	"true"
elasticsearch.roles.master	string	"true"
elasticsearch.roles.remote_cluster_client	string	"true"
elasticsearch.volumeClaimTemplate.accessModes[0]	string	"ReadWriteOnce"
elasticsearch.volumeClaimTemplate.resources.requests.storage	string	"50Gi"
externalServices.elasticsearchHost	string	""	Existing ElasticSearch Hostname to use if elasticsearch.enabled is false
externalServices.elasticsearchPort	int	9200	Existing ElasticSearch Port to use if elasticsearch.enabled is false
externalServices.mongodbHost	string	""	Existing MongoDB Hostname to use if mongodb.enabled is false
externalServices.mongodbPort	int	27017	Existing MongoDB Port to use if mongodb.enabled is false
externalServices.redisHost	string	""	Existing Redis Hostname to use if redis.enabled is false
externalServices.redisPort	int	6379	Existing Redis Port to use if redis.enabled is false
fileserver.affinity	object	{}
fileserver.extraEnvs	list	[]
fileserver.image.pullPolicy	string	"IfNotPresent"
fileserver.image.repository	string	"allegroai/clearml"
fileserver.image.tag	string	"1.7.0"
fileserver.nodeSelector	object	{}
fileserver.podAnnotations	object	{}
fileserver.replicaCount	int	1
fileserver.resources	object	{}
fileserver.service.nodePort	int	30081	If service.type set to NodePort, this will be set to service's nodePort field. If service.type is set to others, this field will be ignored
fileserver.service.port	int	8081
fileserver.service.type	string	"NodePort"	This will set to service's spec.type field
fileserver.storage.data.class	string	""
fileserver.storage.data.size	string	"50Gi"
fileserver.tolerations	list	[]
imageCredentials	object	{"email":"someone@host.com","enabled":false,"existingSecret":"","password":"pwd","registry":"docker.io","username":"someone"}	Private image registry configuration
imageCredentials.email	string	"someone@host.com"	Email
imageCredentials.enabled	bool	false	Use private authentication mode
imageCredentials.existingSecret	string	""	If this is set, chart will not generate a secret but will use what is defined here
imageCredentials.password	string	"pwd"	Registry password
imageCredentials.registry	string	"docker.io"	Registry name
imageCredentials.username	string	"someone"	Registry username
ingress.annotations	object	{}
ingress.api.annotations	object	{}
ingress.api.enabled	bool	false
ingress.api.hostName	string	"api.clearml.127-0-0-1.nip.io"
ingress.api.path	string	"/"
ingress.api.tlsSecretName	string	""
ingress.app.annotations	object	{}
ingress.app.enabled	bool	false
ingress.app.hostName	string	"app.clearml.127-0-0-1.nip.io"
ingress.app.path	string	"/"
ingress.app.tlsSecretName	string	""
ingress.files.annotations	object	{}
ingress.files.enabled	bool	false
ingress.files.hostName	string	"files.clearml.127-0-0-1.nip.io"
ingress.files.path	string	"/"
ingress.files.tlsSecretName	string	""
ingress.name	string	"clearml-server-ingress"
mongodb.architecture	string	"standalone"
mongodb.auth.enabled	bool	false
mongodb.enabled	bool	true
mongodb.persistence.accessModes[0]	string	"ReadWriteOnce"
mongodb.persistence.enabled	bool	true
mongodb.persistence.size	string	"50Gi"
mongodb.replicaCount	int	1
mongodb.service.name	string	"{{ .Release.Name }}-mongodb"
mongodb.service.port	int	27017
mongodb.service.portName	string	"mongo-service"
mongodb.service.type	string	"ClusterIP"
redis.cluster.enabled	bool	false
redis.databaseNumber	int	0
redis.enabled	bool	true
redis.master.name	string	"{{ .Release.Name }}-redis-master"
redis.master.persistence.accessModes[0]	string	"ReadWriteOnce"
redis.master.persistence.enabled	bool	true
redis.master.persistence.size	string	"5Gi"
redis.master.port	int	6379
redis.usePassword	bool	false
secret.authToken	string	"1SCf0ov3Nm544Td2oZ0gXSrsNx5XhMWdVlKz1tOgcx158bD5RV"	Set for auth_token field
secret.credentials.apiserver.accessKey	string	"5442F3443MJMORWZA3ZH"	Set for apiserver_key field
secret.credentials.apiserver.secretKey	string	"BxapIRo9ZINi8x25CRxz8Wdmr2pQjzuWVB4PNASZqCtTyWgWVQ"	Set for apiserver_secret field
secret.credentials.tests.accessKey	string	"ENP39EQM4SLACGD5FXB7"	Set for tests_user_key field
secret.credentials.tests.secretKey	string	"lPcm0imbcBZ8mwgO7tpadutiS3gnJD05x9j7afwXPS35IKbpiQ"	Set for tests_user_secret field
secret.existingSecret	string	""	If this is set, chart will not generate a secret but will use what is defined here
secret.httpSession	string	"9Tw20RbhJ1bLBiHEOWXvhplKGUbTgLzAtwFN2oLQvWwS0uRpD5"	Set for http_session field
webserver.additionalConfigs	object	{}
webserver.affinity	object	{}
webserver.extraEnvs	list	[]
webserver.image.pullPolicy	string	"IfNotPresent"
webserver.image.repository	string	"allegroai/clearml"
webserver.image.tag	string	"1.7.0"
webserver.nodeSelector	object	{}
webserver.podAnnotations	object	{}
webserver.replicaCount	int	1
webserver.resources	object	{}
webserver.service.nodePort	int	30080	If service.type set to NodePort, this will be set to service's nodePort field. If service.type is set to others, this field will be ignored
webserver.service.port	int	80
webserver.service.type	string	"NodePort"	This will set to service's spec.type field
webserver.tolerations	list	[]