mirror of
https://github.com/clearml/clearml-helm-charts
synced 2025-01-31 17:16:47 +00:00
139 lines
4.5 KiB
Go Template
139 lines
4.5 KiB
Go Template
# ClearML Ecosystem for Kubernetes
|
|
{{ template "chart.deprecationWarning" . }}
|
|
|
|
{{ template "chart.badgesSection" . }}
|
|
|
|
{{ template "chart.description" . }}
|
|
|
|
{{ template "chart.homepageLine" . }}
|
|
|
|
{{ template "chart.maintainersSection" . }}
|
|
|
|
## Introduction
|
|
|
|
The **clearml-server** is the backend service infrastructure for [ClearML](https://github.com/allegroai/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
|
|
|
|
## Add to local Helm repository
|
|
|
|
To add this chart to your local Helm repository:
|
|
|
|
```
|
|
helm repo add allegroai https://allegroai.github.io/clearml-helm-charts
|
|
```
|
|
|
|
## Local environment
|
|
|
|
For development/evaluation it's possible to use [kind](https://kind.sigs.k8s.io).
|
|
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.
|
|
|
|
A production ready cluster should also have some different configuration like the one proposed in `values-production.yaml` that can be applied with:
|
|
|
|
```
|
|
helm install clearml allegroai/clearml -f values-production.yaml
|
|
```
|
|
|
|
## 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.
|
|
|
|
### Major upgrade from 5.* to 6.*
|
|
|
|
Before issuing helm upgrade:
|
|
|
|
* delete Redis statefulset(s)
|
|
* scale MongoDB deployment(s) replicas to 0
|
|
* if using securityContexts check for new value form in values.yaml (podSecurityContext and containerSecurityContext)
|
|
|
|
## 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](https://github.com/allegroai/clearml-server#optional-configuration) section in the **clearml-server** repository README file.
|
|
|
|
{{ template "chart.sourcesSection" . }}
|
|
|
|
{{ template "chart.requirementsSection" . }}
|
|
|
|
{{ template "chart.valuesSection" . }}
|