Skip to content

Multi-Container Registration Example

Configure multiple containers with different settings in a single workload.

Tip

Multi-container allows you to register containers A and B — with different images or settings — together in one workload. It is also possible to configure multiple containers using the same image.

What is Multi-Container?

A configuration that registers two or more containers in a single workload. Each container can have independently configured images, ports, environment variables, and commands.

Configuration Container A Container B
Image ollama/ollama:latest open-webui/open-webui:latest
Role LLM inference server Chat UI
Port 11434 8080

Note

Adding more containers does not increase costs — billing is based on the allocated GPU. As long as replicas are not added, the GPU cost ceiling remains unchanged, and all containers share the allocated GPU resources.

Do not confuse Replicas with Multi-Container

In gcube, replicas refer to the number of GPUs deployed in parallel within a single workload. Multi-container is separate — it is a method of configuring multiple containers with different roles within one workload.

Registration Example — Ollama + Open WebUI

An example of configuring Ollama (inference server) and Open WebUI (chat interface) together in a single workload.

① First Container Setup (Ollama)

Item Value
Registry Type Docker Hub
Container Image ollama/ollama:latest
Container Port 11434 (auto-filled)

Screenshot: First container setup

② Add Second Container

  1. Click the Add Container button.

    Screenshot: Add container button

  2. Enter the second container information.

    Item Value
    Registry Type GitHub
    Container Image open-webui/open-webui:ollama
    Container Port 8080 (auto-filled)

    Screenshot: Second container setup

③ Select GPU

Item Value
GPU Model RTX 5090
GPU Memory 32GB

Screenshot: GPU selection

④ Complete Registration

  1. Confirm the estimated total cost.

    Screenshot: Estimated total cost

  2. Select immediate deployment and click the Register button.

Post-Deployment Check

Confirm that all registered containers are running in the Deployment Status tab of the workload details. Screenshot: Deployment status tab — container running confirmation

Warning

If inter-container communication is required (e.g., Open WebUI → Ollama calls), verify port and environment variable settings in advance.