Skip to content
Dataflow

Containers

Containers allow you to deploy and manage containerized applications directly within your Dataflow workspace. Each container deployment creates the necessary infrastructure automatically, making services such as Redis, Nginx, or any custom image instantly accessible from within your development server.

How Containers Work

When you deploy a container, Dataflow provisions it inside the cluster and exposes it via an internal hostname derived from the container’s slug. Your development server can connect to this hostname directly without any additional network configuration.

Deploying a Container

  1. Navigate to the Containers page under Studio settings
  2. Click Deploy Container to open the deployment dialog
  3. Fill in the required details:
    • Name: Display name for your container
    • Slug: Unique identifier (lowercase letters, numbers, and hyphens only)
    • Image: Container image from Docker Hub or a registry (e.g., nginx:latest, redis:7)
    • Port: Port number to expose — optional, range 1–65535
  4. Configure optional fields as needed:
    • Environment Variables: Add variables in KEY=VALUE format, one per line
    • Command: Override the container’s default command (space-separated string)
    • CPU Limit: CPU resource limit in cores (e.g., 1, 2) or millicores (e.g., 500m) — Default: 1, Max: 2
    • Memory Limit: Memory resource limit (e.g., 512Mi, 2Gi) — Default: 2Gi, Max: 8Gi
  5. Click Deploy — the container will be provisioned to the cluster

Container Status

StatusDescription
PENDINGContainer is being created or starting up
DEPLOYEDContainer is running and available
FAILEDContainer deployment failed or is not running
DELETINGContainer is being removed

Click the refresh icon next to the status to fetch the latest real-time status.

Deleting a Container

Click the Delete button next to the target container. This removes the container deployment and all associated resources.

Important Notes

Unique Slugs

Each container must have a unique slug within your organization. The slug is used to generate the internal hostname for the container.

Environment Variables

Must be provided in KEY=VALUE format, one per line. For example:

REDIS_PORT=6379
APP_ENV=production

Resource Limits

  • CPU can be specified in cores (1, 2) or millicores (500m, 1000m)
  • Memory can be specified in Mi (512Mi) or Gi (2Gi, 4Gi)

These limits ensure containers do not consume excessive cluster resources.

Command Override

If specified, the value replaces the container’s default CMD or ENTRYPOINT. Provide the full command as a space-separated string.

Accessing a Container from Your Development Server

Once a container is in the DEPLOYED state, it is reachable from your development server using the internal hostname generated from the slug. The hostname follows the pattern:

<slug>-<unique-id>

Example — Connecting to Redis

After deploying a Redis container (e.g., with slug redis-sc), connect to it from a notebook or script:

import redis


r = redis.Redis(
    host="redis-sc-5a718ce9-aa7a-4b0b-8c9f-2b4c7809b567",
    port=6379,
    decode_responses=True
)


r.set("check", "Working")
print(r.get("check"))  # Output: Working