Skip to main content

Deploying Flagsmith on Google Cloud

Overview​

Flagsmith can be deployed on Google Cloud Platform using either of two approaches:

tip

For production and enterprise deployments, we recommend GKE with our Helm charts. This gives you full control over scaling, networking, and operational tooling.

Cluster Setup​

We recommend GKE Standard or GKE Autopilot depending on how much control you need over node configuration. Either mode works well with Flagsmith.

For node pool sizing guidance, see our sizing and scaling page. As a starting point, our recommended resource allocation per pod is 1 vCPU and 2 GB RAM.

Deploying Flagsmith with Helm​

We publish official Helm charts for deploying Flagsmith on Kubernetes. Add the repository and install:

Install Flagsmith on GKE
helm repo add flagsmith https://flagsmith.github.io/flagsmith-charts/
helm install -n flagsmith --create-namespace flagsmith flagsmith/flagsmith \
  -f values.yaml

A minimal values.yaml for GKE with an external Cloud SQL database looks like this:

values.yaml
postgresql:
 enabled: false

databaseExternal:
 enabled: true
 urlFromExistingSecret:
  enabled: true
  name: flagsmith-database-credentials
  key: DATABASE_URL

api:
 secretKeyFromExistingSecret:
  enabled: true
  name: flagsmith-secret-key
  key: SECRET_KEY
 replicacount: 2

frontend:
 replicacount: 2

taskProcessor:
 enabled: true

Create the Kubernetes secrets separately:

Create secrets
kubectl -n flagsmith create secret generic flagsmith-database-credentials \
  --from-literal=DATABASE_URL="postgres://flagsmith:PASSWORD@CLOUD_SQL_IP:5432/flagsmith"

kubectl -n flagsmith create secret generic flagsmith-secret-key \
  --from-literal=SECRET_KEY="$(openssl rand -hex 32)"
warning

Do not use the in-cluster PostgreSQL for production deployments. Always use an externally managed database such as Cloud SQL for PostgreSQL.

For the full list of chart values — including ingress, PgBouncer, resource limits, and deployment strategy — see our Kubernetes and OpenShift deployment guide.

The chart source is available on GitHub: Flagsmith Helm Charts.

Database — Cloud SQL for PostgreSQL​

We recommend Cloud SQL for PostgreSQL as the database for Flagsmith on GCP.

  • PostgreSQL version: We run our SaaS platform on PostgreSQL 15. Versions 12 and above are supported.
  • High availability: Enable regional HA for production workloads.
  • Connectivity: Use private IP or the Cloud SQL Auth Proxy to connect from GKE.
  • Connection pooling: For high-traffic deployments, enable PgBouncer via the Helm chart. See the PgBouncer section in our Kubernetes guide.

When starting for the first time, Flagsmith will create the database schema automatically. Schema upgrades happen seamlessly during application server upgrades.

Terraform​

info

Terraform modules for provisioning GKE and Cloud SQL for Flagsmith deployments are planned. In the meantime, you can use the Helm chart to deploy Flagsmith onto an existing GKE cluster.

Cloud Run (Quick Start)​

note

Cloud Run is well suited for evaluation and smaller deployments. For production and enterprise workloads, we recommend GKE with Helm charts.

We recommend running the unified Docker image on Cloud Run.

Study our docker-compose file to understand the base environment variables. All available environment variables are documented here.

Run a single Cloud Run service with at least 2 minimum instances to avoid cold starts, particularly for serving low-latency requests to the SDKs. For more sizing information, see our scaling page.

Use /health as the health-check endpoint for both the API and the frontend.

For the database, use Cloud SQL for PostgreSQL. We support PostgreSQL versions 12 and above; our SaaS platform runs on version 15.

tip

When your deployment grows beyond evaluation, consider migrating to GKE with Helm charts for better control over scaling, networking, and operations.