Phase 2 — Pulumi bootstrap

With a bare Nomad node running (Phase 1), the Pulumi stack in the abc-deployments repository installs all abc-cluster services on top: MinIO, tusd (resumable uploads), the monitoring stack, and any supporting Nomad jobs.

Status: The abc-seedling Pulumi stack is in active development. The stack source and detailed usage instructions live in the abc-deployments repository. This page describes what the stack does and the inputs it needs from Phase 1.

What the Pulumi stack installs

Service	Nomad job	Purpose
MinIO	abc-minio	Object storage — one bucket per group
tusd	abc-tusd	Resumable browser uploads (/files/ endpoint)
Grafana	abc-grafana	Dashboards for job and node metrics
Prometheus	abc-prometheus	Metrics collection
MinIO groups + policies	(provisioned via mc)	Per-group ACL bucket isolation

The landing page and claim service (abc-landing-api) are deployed separately in Phase 4 — they depend on the pool database being populated first (Phase 3).

Prerequisites

From Phase 1 you need:

• Nomad management token
• Nomad HTTP address (e.g. http://your-node:4646)
• su-* namespaces created

You also need:

• Pulumi CLI installed on your deploy machine
• Node.js 18+ (for the TypeScript Pulumi program)
• SSH access to the cluster node (for scp and ssh within the stack)

Run the stack

# Clone the abc-deployments repository
git clone https://github.com/abc-cluster/abc-deployments
cd abc-deployments/abc-seedling-stack   # path TBC — check the repo

# Install dependencies
npm install

# Configure the stack
pulumi stack init my-cluster
pulumi config set nomadAddr    http://your-node:4646
pulumi config set nomadToken   <management-token> --secret
pulumi config set clusterHost  your-cluster.example.com
pulumi config set groups       "lab1,lab2,demo"

# Deploy
pulumi up

The stack will:

1. Submit all Nomad jobs (MinIO, tusd, monitoring).
2. Wait for each job to reach running status before proceeding.
3. Run mc commands to create MinIO groups and attach bucket policies.
4. Print the MinIO root credentials and service addresses on completion.

Verify services

After pulumi up completes:

# All abc jobs should be running
nomad job status

# MinIO is healthy
curl http://your-node:9000/minio/health/ready   # 200 OK

# tusd is accepting uploads
curl http://your-node:1080/files/               # 405 Method Not Allowed (correct)

If you configured a hostname and DNS is pointing at the node:

https://minio.your-cluster.example.com/    MinIO console
https://nomad.your-cluster.example.com/    Nomad UI
https://grafana.your-cluster.example.com/  Grafana dashboards

What you have at the end of Phase 2

• Nomad running with ACLs and su-* namespaces.
• MinIO running with per-group bucket policies.
• tusd running for browser uploads.
• Monitoring stack running (optional but included by default).
• MinIO root credentials (needed for Phase 3 provisioning).

Next step

→ Phase 3: Provision the access pool