Skip to main content

Deployment Overview

Flagsmith SaaS Platform

If you would rather skip the hosting and jump straight to integrating Flagsmith with your own application, you can use https://flagsmith.com/ right now. We have paid plans with pricing to suit both startups and enterprise customers alike.

Terraform Templates

We have a number of example deployments across different providers and orchestration frameworks in our Terraform Examples repository.

One Click Installers

Deploy to DigitalOcean

Deploy to Render

Deploy on Railway

Fly.io

We're big fans of Fly.io! You can deploy to fly.io really easily:

git clone git@github.com:Flagsmith/flagsmith.git
cd flagsmith
flyctl postgres create --name flagsmith-flyio-db
flyctl apps create flagsmith-flyio
flyctl postgres attach --postgres-app flagsmith-flyio-db
flyctl deploy

Fly.io has a global application namespace, and so you may need to change the name of the application defined in fly.toml as well as the commands above.

Caprover

You can also deploy to a Caprover Server with One Click Apps.

Self Hosting Overview

You will need to run through the following steps to get set up:

  1. Create a Postgres database to store the Flagsmith data.
  2. Deploy the API and set up DNS for it. If you are using health-checks, make sure to use /health as the health-check endpoint.
  3. Visit http://<your-server-domain:8000>/api/v1/users/config/init/ to create an initial Superuser and provide DNS info to the platform.
  4. Deploy the Front End Dashboard and set up DNS for it. Point the Dashboard to the API using the relevant Environment Variables. If you are using health-checks, make sure to use /health as the health-check endpoint.
  5. Create a new Organisation, Project, Environment and Flags via the Dashboard.
  6. When using our SDKs, you will need to override the API URL that they point to, otherwise they will default to connect to our paid-for API at https://edge.api.flagsmith.com/api/v1. See the SDK documentation for the library you are using.

Deployment Options

We recommend running Flagsmith with Docker. We have options to run within Docker, Kubernetes or RedHat OpenShift.

Architecture

The Flagsmith architecture is based around a REST API that is accessed by both SDK clients and the Flagsmith Dashboard Front End Web App.

Application Architecture

Dependencies

Running the API has the following hard dependencies:

  • Postgres database - the main data store. We have tested and run against Postgres v11.12 but it can work for other versions too.

The API can also optionally make use of the following 3rd party services:

  • Google Analytics - for API analytics
  • InfluxDB - for API analytics
  • SendGrid - for transactional email
  • AWS S3 - to store Django Static Assets
  • GitHub - oAuth provider
  • Google - oAuth provider

Flag Analytics

Flagsmith stores time series data for two use cases:

  1. Flag Analytics
  2. API traffic reporting

Flagsmith can be configured to store and process this data in one of three ways:

  1. To store it in Postgres
  2. To store it in InfluxDB
  3. To not store it at all

We recommend option 1.

Time series data via Postgres

Add the following environment variables to the Flagsmith API service:

# Set Postgres to store the data
USE_POSTGRES_FOR_ANALYTICS=True

# Configure the postgres datastore:
# Either
ANALYTICS_DATABASE_URL (e.g. postgresql://postgres:password@postgres:5432/flagsmith)
# Or
DJANGO_DB_HOST_ANALYTICS (e.g. postgres.db)
DJANGO_DB_NAME_ANALYTICS (e.g. flagsmith)
DJANGO_DB_USER_ANALYTICS (e.g. postgres_user)
DJANGO_DB_PASSWORD_ANALYTICS (e.g. postgres_password)
DJANGO_DB_PORT_ANALYTICS (e.g. 5432)

Note that you don't have to use the same database or database server as the core Flagsmith DB.

You will also need to be running the Task Processor for downsampling to work and the stats to start showing up in the dashboard. This process can take up to 1 hour.

API Telemetry

Flagsmith collects information about self hosted installations. This helps us understand how the platform is being used. This data is never shared outside of the organisation, and is anonymous by design. You can opt out of sending this telemetry on startup by setting the ENABLE_TELEMETRY environment variable to False.

We collect the following data on startup and then once every 8 hours per API server instance:

  • Total number of Organisations
  • Total number of Projects
  • Total number of Environments
  • Total number of Features
  • Total number of Segments
  • Total number of Users
  • DEBUG django variable
  • ENV django variable
  • API server external IP address

Running Flagsmith on Flagsmith

Flagsmith uses Flagsmith to control features on the front end dashboard. If you are self hosting the platform, you will sometimes see features greyed out, or you may want to disable specific features, e.g. logging in via Google and Github. If you are using your own Flagsmith environment then you will need to have a replica of our flags in order to control access to those features.

To do this,firstly create a new project within your self-hosted Flagsmith application. This is the project that we will use to control the features of the self-hosted Flagsmith instance. We will then point the self hosted front end dashboard at this Flagsmith project in order to control what features show for your self hosted Flagsmith instance.

Once you have created the project, you need to set the following Front End environment variables in order to configure this:

  • FLAGSMITH_ON_FLAGSMITH_API_KEY
    • The Flagsmith Client-side Environment Key we use to manage features - Flagsmith runs on Flagsmith. This will be the API key for the project you created as instructed above.
  • FLAGSMITH_ON_FLAGSMITH_API_URL
    • The API URL which the Flagsmith front end dashboard should communicate with. This will most likely be the domain name of the Flagsmith API you are self hosting: Flagsmith runs on Flagsmith. E.g. For our SaaS hosted platform, the variable is https://edge.api.flagsmith.com/api/v1/. For example, if you were running everything locally using the standard docker-compose setup, you would use http://localhost:8000/api/v1/

Once you have set this up, you should see the Flagsmith front end requesting its own flags from the API (you can look in your browser developer console to see this). You can now start creating flags and overriding the default behaviours of the platform. For example, if you wanted to disable Google OAuth authentication, you would create a flag called oauth_google and disable it.

Current Flagsmith Feature Flags

The list of the flags and remote config we're currently using in production is below:

Flag NameDescriptionText Value
4eyesWhether to enable the Change Requests featureNone
butter_barmarkdown to show at the top of the dashboard pageNone
case_sensitive_flagsEnables the project setting to allow case sensitive flagsNone
compare_environmentsCompare feature flag changes across environmentsNone
dark_modeEnables Dark Mode in UI See BelowNone
disable_create_orgTurning this on will prevent users from creating any additional organisationsNone
feature_name_regexEnables the project setting to add a regex matcher to validate feature namesNone
flag_analyticsFlag usage chart - requires InfluxDBNone
force_2faEnables the organisation setting to force 2 factor authenticationNone
integration_dataConfigures integrationsSee Below
integrationsWhich third party integrations are displayed["amplitude","datadog","dynatrace","heap","mixpanel","new-relic","rudderstack","segment","slack","webhook"]
oauth_githubConfigure Github OAuthSee Below
oauth_googleConfigure Google OAuthSee Below
rotate_api_tokenEnables the ability to rotate a user's access tokenSee Below
samlEnables SAML authenticationSee Below
scaleup_auditDisables audit log for anyone under scale-up planNone
segment_associated_featuresEnables the ability to see features associated with a segmentNone
segment_operatorsDetermines what rules are shown when creating a segmentSee Below
serverside_sdk_keysEnable Server-side Environment KeysNone
sso_idpSet this to your configured SAML organisation name to automatically redirect to your IdPNone
tag_environmentsEnables an environment setting to add a UI hint to your environments (e.g. for prod)None
try_itWhether to show the try it buttonsNone
usage_chartOrganisation Analytics usage chart - requires InfluxDBNone

integration_data

{
"datadog": {
"perEnvironment": false,
"image": "/static/images/integrations/datadog.svg",
"docs": "https://docs.flagsmith.com/integrations/apm/datadog/",
"fields": [
{
"key": "base_url",
"label": "Base URL"
},
{
"key": "api_key",
"label": "API Key"
}
],
"tags": ["logging"],
"title": "Datadog",
"description": "Sends events to Datadog for when flags are created, updated and removed. Logs are tagged with the environment they came from e.g. production."
},
"dynatrace": {
"perEnvironment": true,
"image": "/static/images/integrations/dynatrace.svg",
"docs": "https://docs.flagsmith.com/integrations/apm/dynatrace/",
"fields": [
{
"key": "base_url",
"label": "Base URL"
},
{
"key": "api_key",
"label": "API Key"
},
{
"key": "entity_selector",
"label": "Entity Selector"
}
],
"tags": ["logging"],
"title": "Dynatrace",
"description": "Sends events to Dynatrace for when flags are created, updated and removed. Logs are tagged with the environment they came from e.g. production."
},
"slack": {
"perEnvironment": true,
"isOauth": true,
"image": "/static/images/integrations/slack.svg",
"docs": "https://docs.flagsmith.com/integrations/slack/",
"tags": ["messaging"],
"title": "Slack",
"description": "Sends messages to Slack when flags are created, updated and removed. Logs are tagged with the environment they came from e.g. production."
},
"amplitude": {
"perEnvironment": true,
"image": "/static/images/integrations/amplitude.svg",
"docs": "https://docs.flagsmith.com/integrations/analytics/amplitude/",
"fields": [
{
"key": "api_key",
"label": "API Key"
}
],
"tags": ["analytics"],
"title": "Amplitude",
"description": "Sends data on what flags served to each identity."
},
"new-relic": {
"perEnvironment": false,
"image": "/static/images/integrations/new_relic.svg",
"docs": "https://docs.flagsmith.com/integrations/apm/newrelic",
"fields": [
{
"key": "base_url",
"label": "New Relic Base URL"
},
{
"key": "api_key",
"label": "New Relic API Key"
},
{
"key": "app_id",
"label": "New Relic Application ID"
}
],
"tags": ["analytics"],
"title": "New Relic",
"description": "Sends events to New Relic for when flags are created, updated and removed."
},
"segment": {
"perEnvironment": true,
"image": "/static/images/integrations/segment.svg",
"docs": "https://docs.flagsmith.com/integrations/analytics/segment",
"fields": [
{
"key": "api_key",
"label": "API Key"
}
],
"tags": ["analytics"],
"title": "Segment",
"description": "Sends data on what flags served to each identity."
},
"rudderstack": {
"perEnvironment": true,
"image": "/static/images/integrations/rudderstack.svg",
"docs": "https://docs.flagsmith.com/integrations/analytics/rudderstack",
"fields": [
{
"key": "base_url",
"label": "Rudderstack Data Plane URL"
},
{
"key": "api_key",
"label": "API Key"
}
],
"tags": ["analytics"],
"title": "Rudderstack",
"description": "Sends data on what flags served to each identity."
},
"webhook": {
"perEnvironment": true,
"image": "/static/images/integrations/webhooks.svg",
"docs": "https://docs.flagsmith.com/integrations/webhook",
"fields": [
{
"key": "url",
"label": "Your Webhook URL Endpoint"
},
{
"key": "secret",
"label": "Your Webhook Secret"
}
],
"tags": ["analytics"],
"title": "Webhook",
"description": "Sends data on what flags served to each identity to a Webhook Endpoint you provide."
},
"heap": {
"perEnvironment": true,
"image": "/static/images/integrations/heap.svg",
"docs": "https://docs.flagsmith.com/integrations/analytics/heap",
"fields": [
{
"key": "api_key",
"label": "API Key"
}
],
"tags": ["analytics"],
"title": "Heap Analytics",
"description": "Sends data on what flags served to each identity."
},
"mixpanel": {
"perEnvironment": true,
"image": "/static/images/integrations/mp.svg",
"docs": "https://docs.flagsmith.com/integrations/analytics/mixpanel",
"fields": [
{
"key": "api_key",
"label": "Project Token"
}
],
"tags": ["analytics"],
"title": "Mixpanel",
"description": "Sends data on what flags served to each identity."
}
}

segment_operators

[
{
"value": "EQUAL",
"label": "Exactly Matches (=)"
},
{
"value": "NOT_EQUAL",
"label": "Does not match (!=)"
},
{
"value": "PERCENTAGE_SPLIT",
"label": "% Split"
},
{
"value": "GREATER_THAN",
"label": ">"
},
{
"value": "GREATER_THAN_INCLUSIVE",
"label": ">="
},
{
"value": "LESS_THAN",
"label": "<"
},
{
"value": "LESS_THAN_INCLUSIVE",
"label": "<="
},
{
"value": "GREATER_THAN:semver",
"label": "SemVer >",
"append": ":semver"
},
{
"value": "GREATER_THAN_INCLUSIVE:semver",
"label": "SemVer >=",
"append": ":semver"
},
{
"value": "LESS_THAN:semver",
"label": "SemVer <",
"append": ":semver"
},
{
"value": "LESS_THAN_INCLUSIVE:semver",
"label": "SemVer <=",
"append": ":semver"
},

{
"value": "CONTAINS",
"label": "Contains"
},
{
"value": "NOT_CONTAINS",
"label": "Does not contain"
},
{
"value": "REGEX",
"label": "Matches regex"
},
{
"value": "MODULO",
"label": "Modulo operation"
},
{
"value": "IS_SET",
"label": "Is set",
"hideValue": true
},
{
"value": "IS_NOT_SET",
"label": "Is not set",
"hideValue": true
}
]

oauth_github

Find instructions for GitHub Authentication here.

Create an OAuth application in the GitHub Developer Console and then provide the following as the Flag value:

  • Create the Flagsmith on Flagsmith flag as below replacing your client_id and redirect_uri
{
"url": "https://github.com/login/oauth/authorize?scope=user&client_id=<your client_id>&redirect_uri=<your url encoded redirect uri>"
}

For example, our SaaS value looks like this (but with our Client ID redacted)

{
"url": "https://github.com/login/oauth/authorize?scope=user&client_id=999999999999&redirect_uri=https%3A%2F%2Fapp.flagsmith.com%2Foauth%2Fgithub"
}

oauth_google

Create an OAuth application in the Google Developer Console and then provide the following as the Flag value:

{
"clientId": "<Your Google oAuth Client ID>",
"apiKey": "<Your Google oAuth Client secret>"
}

Dark Mode

We also have a Segment that manages the ui Dark Mode:

Segment Name: dark_mode Segment Rules: Trait dark_mode EXACTLY MATCHES True

Then use this rule to override the dark_mode Feature Flag.

Integrations

Some Integrations require additional configuration

Slack Integration

Create a private Slack app. You will then need to provide the following environment variables on the API side:

  • SLACK_CLIENT_ID
  • SLACK_CLIENT_SECRET

You can retrieve these values from Slack. You will also need to add the following scopes:

  • channels:read
  • chat:write
  • chat:write.public

You will also need to set up the redirect URLs for your application. For more information on this see Slack's docs on creating your own app, and the OAuth flow that goes along with that. The production Flagsmith App Manifest reads as follows and can be used as a template:

{
"display_information": {
"name": "Flagsmith Bot",
"description": "Get notified in Slack whenever changes are made to your Flagsmith Environments",
"background_color": "#000000",
"long_description": "Use our application for Slack to receive Flagsmith state changes directly in your Slack channels. Whenever you create, update or delete a Flag within Flagsmith, our application for Slack will send a message into a Slack channel of your choosing.\r\n\r\nFlagsmith is an open source, fully featured, Feature Flag and Remote Config service. Use our hosted API, deploy to your own private cloud, or run on-premise."
},
"features": {
"bot_user": {
"display_name": "Flagsmith Bot",
"always_online": false
}
},
"oauth_config": {
"redirect_urls": [
"https://api.flagsmith.com/api/v1/environments",
"https://api-staging.flagsmith.com/api/v1/environments"
],
"scopes": {
"bot": ["channels:read", "chat:write", "chat:write.public"]
}
},
"settings": {
"org_deploy_enabled": false,
"socket_mode_enabled": false,
"token_rotation_enabled": false
}
}

Time series data via InfluxDB

Flagsmith has a soft dependency on InfluxDB to store time-series data. You don't need to configure Influx to run the platform; by default this data will be stored in Postgres. If you are running very high traffic loads, you might be interested in deploying InfluxDB.

  1. Create a user account in influxdb. You can visit http://localhost:8086/
  2. Go into Data > Buckets and create three new buckets called default, default_downsampled_15m and default_downsampled_1h
  3. Go into Data > Tokens and grab your access token.
  4. Edit the docker-compose.yml file and add the following environment variables in the api service to connect the api to InfluxDB:
    • INFLUXDB_TOKEN: The token from the step above
    • INFLUXDB_URL: http://influxdb
    • INFLUXDB_ORG: The organisation ID - you can find it here
    • INFLUXDB_BUCKET: default
  5. Restart docker-compose
  6. Create a new task with the following query. This will downsample your per millisecond api request data down to 15 minute blocks for faster queries. Set it to run every 15 minutes.
option task = {name: "Downsample (API Requests)", every: 15m}

data = from(bucket: "default")
|> range(start: -duration(v: int(v: task.every) * 2))
|> filter(fn: (r) =>
(r._measurement == "api_call"))

data
|> aggregateWindow(fn: sum, every: 15m)
|> filter(fn: (r) =>
(exists r._value))
|> to(bucket: "default_downsampled_15m")

Once this task has run you will see data coming into the Organisation API Usage area.

  1. Create another new task with the following query. This will downsample your per millisecond flag evaluation data down to 15 minute blocks for faster queries. Set it to run every 15 minutes.
option task = {name: "Downsample (Flag Evaluations)", every: 15m}

data = from(bucket: "default")
|> range(start: -duration(v: int(v: task.every) * 2))
|> filter(fn: (r) =>
(r._measurement == "feature_evaluation"))

data
|> aggregateWindow(fn: sum, every: 15m)
|> filter(fn: (r) =>
(exists r._value))
|> to(bucket: "default_downsampled_15m")

Once this task has run, and you have made some flag evaluations with analytics enabled (see documentation here for information on this) you should see data in the 'Analytics' tab against each feature in your dashboard.

  1. Create another new task with the following query. This will downsample your per millisecond api request data down to 1 hour blocks for faster queries. Set it to run every 1 hour.
option task = {name: "Downsample API 1h", every: 1h}

data = from(bucket: "default")
|> range(start: -duration(v: int(v: task.every) * 2))
|> filter(fn: (r) =>
(r._measurement == "api_call"))

data
|> aggregateWindow(fn: sum, every: 1h)
|> filter(fn: (r) =>
(exists r._value))
|> to(bucket: "default_downsampled_1h")
  1. Create another new task with the following query. This will downsample your per millisecond flag evaluation data down to 1 hour blocks for faster queries. Set it to run every 1 hour.
option task = {name: "Downsample API 1h - Flag Analytics", every: 1h}

data = from(bucket: "default")
|> range(start: -duration(v: int(v: task.every) * 2))
|> filter(fn: (r) =>
(r._measurement == "feature_evaluation"))
|> filter(fn: (r) =>
(r._field == "request_count"))
|> group(columns: ["feature_id", "environment_id"])

data
|> aggregateWindow(fn: sum, every: 1h)
|> filter(fn: (r) =>
(exists r._value))
|> set(key: "_measurement", value: "feature_evaluation")
|> set(key: "_field", value: "request_count")
|> to(bucket: "default_downsampled_1h")

Manual Installation

If you want a more configurable environment, you can manually install both the Front End and the API.

Server Side API

The source code and installation instructions can be found at the GitHub project. The API is written in Python and is based on Django and the Django Rest Framework. The Server side API relies on a Postgres SQL installation to store its data, and a Redis installation as a cache.

Front End Website

The source code and installation instructions can be found at the GitHub project. The Front End Website is written in React/Javascript and requires NodeJS.