Attention: Deprecation notice for Bintray, JCenter, GoCenter and ChartCenter. Learn More


Chart version: 1.13.0
Api version: v1
App version: 5.29.0
Mattermost Enterprise server with high availitibity.
Chart Type
Set me up:
helm repo add center
Install Chart:
helm install mattermost-enterprise-edition center/mattermost/mattermost-enterprise-edition
Versions (0)

Mattermost Enterprise Edition Helm Chart

This is the Helm chart for the Mattermost Enterprise Edition. It is subject to changes, but is used by Mattermost internally to run CI servers and our server. To learn more about Helm charts, see the Helm docs.

To read an overview of how to migrate Mattermost to Kubernetes, see this blog post.

The Mattermost Team Edition Helm chart can be found here.

To install the Mattermost Team Edition Helm chart in a GitLab Helm chart deployment, see this documentation.

1. Prerequisites

1.1 Kubernetes Cluster

You need a running Kubernetes cluster v1.8+. If you do not have one, find options and installation instructions here:

1.2 Helm


We recommend installing Helm v2.13.1 or later.

Once Helm is installed and initialized, run the following:

helm repo add mattermost
helm repo add incubator

1.3 Ingress

To expose the application outside of your network, configure an ingress if your Kubernetes cluster doesn’t already have one.

We suggest using nginx-ingress, which we use internally at Mattermost.

To install nginx-ingress in your Kubernetes cluster, follow the official installation documentation. You may also use the helm charts directly.

To get the nginx cache to work, adjust the ConfigMap and Deployment from the above instructions. See the ingress section of our blog post to see how we did it for an AWS installation.

1.4 Certificate Manager

If you do not want to manually add SSL/TLS certificates, install cert-manager. You can follow this documentation or install the helm charts.

To run with HTTPS you will need to add a Kubernetes secret for your SSL/TLS certificate, whether you use cert-manager or not.

2. Configuration

To start, copy mattermost-enterprise-edition/values.yaml and name it config.yaml. This will be your configuration file for the Mattermost Helm chart.

To learn more about the Mattermost configuration file, see

2.1 Required Settings

At minimum there are two settings you must update:

  • global.siteURL - set this to the URL your users will use to access Mattermost, e.g.
  • global.mattermostLicense - set this to the contents of your license file

Without these two settings, Mattermost will not run correctly.

2.2 Application Version

To set the Mattermost application version, update:

  • mattermostApp.image.tag - set this to the Mattermost server version you wish to install (e.g. 5.12.0)

2.3 Ingress

If you are using nginx-ingress, set the following under mattermostApp:

  enabled: true

where is your domain name and matches global.siteURL.

2.3.1 HTTPS

To run with HTTPS, add an SSL/TLS certificate as a secret to your Kubernetes cluster, either manually or using cert-manager.

Set the following under mattermostApp to enable HTTPS:

  enabled: true
    - secretName: your-tls-secret-name

2.3.2 DNS

To route users from your domain name to your Mattermost installation, point your domain name at the external IP or domain that your ingress exposes.

Depending on the DNS service and Ingress you’re using, the steps can vary. If you are using nginx-ingress, you would do something like this:

  1. Run kubectl describe svc your-nginx-ingress-controller
    • Replace your-nginx-ingress-controller with the name of your ingress controller service
  2. Copy the domain name beside LoadBalancer Ingress:
  3. On your DNS service, create a CNAME record pointing from the domain you’d like to use to the domain name you just copied
  4. Save, and wait 10-15 minutes for the DNS change to propagate

2.4 Database

For database configuration, you have two options:

  1. Use an internal MySQL database managed by the Mattermost Helm chart
  2. Use an external database not managed by the Mattermost Helm chart

By default the chart uses the internal MySQL database.

2.4.1 Internal

We use incubator/mysqlha to run the internal database.

The internal database is configured by default, but we recommend updating the following settings:

  • mysqlha.mysqlha.mysqlRootPassword
  • mysqlha.mysqlha.mysqlUser
  • mysqlha.mysqlha.mysqlPassword

See the incubator/mysqlha configuration settings for more options you can configure.

To backup / restore your database, please follow this how-to.

2.4.2 External

If you would like to use an external database not managed by the Mattermost Helm chart, do the following:

  • Set mysqlha.enabled to false
  • Set global.features.database.useInternal to false
  • Set global.features.database.external.driver to either mysql or postgres
  • Set global.features.database.external.dataSource to your master DB connection string
  • (Optional) Set global.features.database.external.dataSourceReplicas to an array of read replica connection strings

2.5 Push Notifications

By default push notifications are enabled using HPNS. If you prefer to run your own push proxy, do the following:

  • Set global.features.notifications.useHPNS to false
  • Install the push proxy Helm chart, see instructions here

2.6 Storage

We use stable/minio for file storage.

Minio is configured by default, but we recommend updating the following settings:

  • minio.accessKey
  • minio.secretKey

See the stable/minio configuration settings for more options you can configure.

3. Install

After adding the Mattermost repo (see section 1.2) you can install a version of the preferred chart by running:

$ helm repo add mattermost
$ helm install mattermost/mattermost-enterprise-edition --version <version_number>

For example:

$ helm repo add mattermost
$ helm install mattermost/mattermost-enterprise-edition --version v0.8.2

If no Helm Chart version is specified the latest version will be installed.

To run with your custom config.yaml, install using:

$ helm install -f config.yaml mattermost/mattermost-enterprise-edition

To upgrade an existing release, modify the config.yaml with your desired changes and then use:

$ helm upgrade -f config.yaml <your-release-name> mattermost/mattermost-enterprise-edition

3.1 Uninstalling Mattermost Enterprise Helm Chart

If you are done with your deployment and want to delete it, use helm delete <your-release-name>. If you don’t know the name of your release, use helm ls to find it.

3.2 Install Push Proxy

You can launch the Mattermost push proxy chart with:

$ helm repo add mattermost
$ helm install mattermost/mattermost-push-proxy

To list options for mattermost-push-proxy:

$ helm inspect values mattermost/mattermost-push-proxy

4. Upgrading to Mattermost Chart Version 1.0.0

This chart version uses the configuration in the database and environment variables for non-changing configuration.

To upgrade from a previous chart version to 1.0.0 you need to follow theses steps before applying the new chart:

  • Get the existing configuration from your Mattermost instance. For that you can use the CLI command. You need to connect to the running pod and run the Mattermost CLI to get the existing config
$ kubectl exec -it -n <YOUR_MATTERMOST_NAMESPACE> <MATTERMOST_APP_POD> -- /bin/sh

in the pod:
$ ./bin/mattermost config show --json
  • Save this config in your local file system
  • Perform the Helm Chart upgrade
  • Copy the config you saved before to the new pod
  • Perform the config migration
$ ./bin/mattermost config migrate <CONFIG_THAT_YOU_COPY_TO_THE_POD.json> $MM_CONFIG
{"level":"info","msg":"Successfully migrated config."}
  • Kill the pods to apply the configs

5. Developing

If you are going to modify the helm charts, it is helpful to use --dry-run (doesn’t do an actual deployment) and --debug (print the generated config files) when running helm install.

Helm has partial support for pulling values out of a subchart via the requirements.yaml. It also has limited support for pushing values into subcharts. It does not support using templating inside a values.yaml file.

We recommend using kind for local development if you do not have access to Kubernetes cluster running in the cloud.