Examples

This page provides step-by-step examples for deploying common workloads on Anvil's Composable Subsystem.

---

Interactive Tutorials:

• Deploying a Kubernetes cluster on Anvil Composable

Database Deployment (PostGIS)

This example demonstrates how to deploy a PostGIS database and expose it to external clients.

Deploy a PostGIS Database

1. Project Selection: Choose your project from the top-right dropdown menu

2. Workload Creation: Navigate to the Workload section via the left menu and click Create

3. Deployment Configuration: Select "Deployment" as your deployment type

4. Form Setup:

   • Select your namespace
   • Assign an arbitrary name to the deployment
   • Set the container image to: registry.anvil.rcac.purdue.edu/docker-hub-cache/postgis/postgis:latest

5. Environment Configuration:

   • Add environment variable via the "Add Variable" button
   • Set POSTGRES_PASSWORD with your chosen password

6. Storage Setup:

   • Access the Storage tab
   • Create a persistent volume claim with single-node read/write access
   • Select appropriate storage class and capacity (example: 5 GiB)
   • Mount point: /var/lib/postgresql/data
   • Set sub-path to data

7. Resource Allocation:

   • CPU Reservation: 2000 (ensures 2 CPU cores available)
   • CPU Limit: 2000 (prevents exceeding 2 CPU cores)

8. Pod Labeling:

   • Add labels under "Labels & Annotations"
   • Example: Key my-db, Value postgis

9. Deployment Launch: Click Create to initiate the deployment

Note

Deployment typically completes within 5 minutes.

Expose the Database Externally

Create a LoadBalancer service for external client access:

1. Service Creation: Navigate to Service Discovery > Services and select Create

2. LoadBalancer Selection: Choose Load Balancer service type

3. Configuration:

   • Select correct namespace
   • Provide service name (results in DNS: <servicename>.<namespace>.anvilcloud.rcac.purdue.edu)
   • Listening Port and Target Port: 5432 (PostgreSQL default)

4. Selector Configuration:

   • Match the pod label from deployment step
   • The interface shows green confirmation when labels match correctly

5. Network Configuration:

   • Private network: Add annotation with Key metallb.universe.tf/address-pool and Value anvil-private-pool
   • Public network: Use Value anvil-public-pool

Verification

Confirm service deployment:

kubectl -n <namespace> get services

Verify DNS record creation:

host <servicename>.<namespace>.anvilcloud.rcac.purdue.edu

The system automatically assigns an IP address and configures DNS naming.

---

Web Server Deployment (Nginx)

This example covers deploying an Nginx web server and exposing it via an Ingress for external access.

Deploy Nginx

1. Select Project: Choose your project from the top-right dropdown

2. Navigate to Workloads: Use the left menu to access the Workload section

3. Create New Workload: Click Create in the top right

4. Choose Deployment Type: Select "Deployment"

5. Configure the Deployment:

   • Select your namespace
   • Assign an arbitrary name for the deployment
   • Set container image to: registry.anvil.rcac.purdue.edu/docker-hub-cache/library/nginx

6. Service Configuration:

   • Click "Add Port"
   • Select "Cluster IP" as the Service Type
   • Set Private Container Port to 80

7. Pod Labels Setup:

   • Navigate to "Labels & Annotations" in the form
   • Select "Add Label" under Pod Labels section
   • Assign a memorable key-value pair (example: Key: my-web, Value: nginx)
   • Click Create

Note

Deployment typically completes within 5 minutes. Initial "does not have minimum availability" messages are normal during deployment startup.

Expose via Ingress

1. Navigate to Service Discovery: Access "Service Discovery" > "Ingresses"

2. Create Ingress: Click Create in the top right

3. Configure Ingress Settings:

   • Select the namespace containing your nginx deployment
   • Provide an arbitrary name
   • Set Request Host (example: my-nginx.anvilcloud.rcac.purdue.edu)
   • Set Path Prefix to /
   • Select the target service and port from dropdowns
   • Click Create

Troubleshooting

Access deployment logs by clicking the workload name, then selecting "View Logs" via the ellipsis menu on the deployed pod.