Deploy a workload - IDYL Docs

Use idyl deploy when you want to create a workload from a container image.

1. Confirm your target

idyl whoami

Check the account, namespace, and subnet. To use a different subnet for one command, pass --subnet.

2. Deploy a service

Use a deployment for software that should keep running:

idyl deploy nginx:1.25 --name web --replicas 2

Output similar to:

✓ Deployment web created
  2 replicas · nginx:1.25 · subnet <subnet>

3. Inspect the deployment

idyl get deployment web
idyl get pod

To watch pods as they change:

idyl get pod --watch

Deploy a job

Use a job for work that should run to completion:

idyl deploy busybox:latest --name batch --completions 1 --restart on_failure

Check the job and pods:

idyl get job batch
idyl get pod --all

Add resources or environment variables

idyl deploy myapp:v2 --name api --cpu 500m --memory 256Mi --env LOG_LEVEL=debug

Use GPU requirements when the subnet has GPU capacity:

idyl deploy vllm:latest --name llm --gpu h100

Request workload isolation

When the target subnet allows it, request microVM-backed execution with --isolation-class microvm:

idyl deploy nginx:1.27-alpine \
  --name web \
  --subnet <subnet> \
  --platform linux/amd64 \
  --isolation-class microvm

The subnet runtime policy and compatible provider capacity must admit the requested isolation class. For the builder, operator, and provider model, see Workload isolation.

Change the target subnet

For a single deployment:

idyl deploy nginx:1.25 --name web --subnet <subnet>

For future commands in the current context:

idyl context edit --set-subnet <subnet>

​1. Confirm your target

​2. Deploy a service

​3. Inspect the deployment

​Deploy a job

​Add resources or environment variables

​Request workload isolation

​Change the target subnet