Prerequisites

This section is written for readers who are new to Kubernetes.

What you are installing

The Flow-Like Kubernetes backend consists of:

• API service: a long-running Kubernetes Deployment.
• Executor: short-lived Kubernetes Jobs created by the API to run workflow executions.
• Optional dependencies (“auto mode”): internal PostgreSQL if you don’t provide external credentials.
• Required external service: S3-compatible storage (AWS S3, Cloudflare R2, Google Cloud Storage, etc.)

Kubernetes basics (quick glossary)

• Cluster: the Kubernetes environment you deploy into.
• Namespace: a logical partition inside a cluster. We recommend a dedicated namespace like flow-like.
• Pod: the unit that runs containers.
• Deployment: keeps a set of pods running (used for the API).
• Job: runs pods until completion (used for workflow executions).
• Service: stable network endpoint inside the cluster.
• Secret: stores sensitive values like database credentials.

Required tools (on your machine)

• kubectl (to talk to your cluster)
• helm (to install the chart)

Required cluster capabilities

• A working Kubernetes cluster (managed Kubernetes, k3s, MicroK8s, etc.)
• A default StorageClass (recommended)
  • Needed if you want persistent volumes for internal Postgres/Redis.
• Outbound network access from the API/executor pods to:
  • your database (if external)
  • your S3 endpoint (if external)

Optional (but common) add-ons

• Ingress controller (NGINX, Traefik, etc.)
  • Needed if you want to expose the API outside the cluster.
• DNS + TLS (cert-manager)
  • Needed if you want HTTPS with a real domain.

Optional security/runtime features

Kata containers

The executor jobs can use a RuntimeClass (for example Kata) for stronger isolation.

• If your cluster does not support Kata, you can still run the executor jobs without it.
• If you do use Kata, your cluster must already have the runtime installed and configured.

Credentials you may need

You can run everything “internally” (auto mode), or you can bring your own external services.

External PostgreSQL

If you use an external database, you’ll typically provide:

• a PostgreSQL connection string (DATABASE_URL) via a Kubernetes secret

S3-compatible storage (required)

Flow-Like requires external S3-compatible storage. You’ll need to provide:

• endpoint
• region
• access key id / secret access key
• bucket names (metadata + content)