Quick Start Guide¶

Get up and running with Kraken Applications in minutes! This guide will walk you through creating and deploying your first manifest.

Prerequisites¶

Before you begin, ensure you have:

• Access to a Scale Computing HyperCore environment
• Kraken Fleet Manager UI or API access
• Basic understanding of YAML syntax

Step 1: Create Your First Application Manifest¶

Let's create a simple virtual machine manifest.

1. Open the "Deployments" page of fleet.scalecomputing.com
2. Select "Applications"
3. Add an application titled "My First VM"
4. Copy-paste the following into the manifest code editor

type: Application
version: "1.0.0"
metadata:
  name: "my-first-vm"
  labels:
    - "environment:development"
    - "purpose:learning"
spec:
  assets:
    - name: "alpine-virt-iso"
      type: "virtual_disk"
      format: "iso"
      url: "https://dl-cdn.alpinelinux.org/alpine/v3.22/releases/x86_64/alpine-virt-3.22.2-x86_64.iso" 
  resources:
    - type: virdomain
      name: "quickstart-vm {{clusterName}}"
      spec:
        description: "My first Kraken VM"
        cpu: 1
        memory: "2 GiB"
        machine_type: "bios"
        state: "running"
        storage_devices:
          - name: "alpine-iso"
            type: "ide_cdrom"
            source: "alpine-virt-iso"
            boot: 2
          - name: "storage_disk"
            type: "virtio_disk"
            capacity: "10 GB"
            boot: 1
        network_devices:
          - name: "eth0"
            type: "virtio"
        tags:
          - "quickstart"
          - "tutorial"

What This Manifest Does¶

• Creates a VM named "quickstart-vm [your cluster name]"
• Allocates resources: 1 CPU core, 2 GiB RAM
• Configures storage: 10 GB VirtIO disk
• Sets up networking: Single VirtIO network interface
• Starts the VM: state: "running"

Step 2: Select Your Target Clusters¶

1. Open the Deployments page in Fleet Manager
2. Navigate to the Cluster Groups section
3. Click "Create Cluster Group"
4. Select a desired test cluster or two
5. Click "Save

Step 3: Create and run Deployment¶

1. Open the Deployments page in Fleet Manager
2. Click "Add Deployment"
3. Name your deployment
4. Select the cluster group you created
5. Attach the application you created
6. Click "Save
7. Click Deploy to send your application(s) to the cluster(s) in your cluster group

Step 4: Monitor Deployment¶

1. Check the Deployments dashboard
2. Monitor deployment progress
3. View resource status and logs

Step 5: Verify Your VM¶

Once deployed, verify your VM is running:

1. Check HyperCore console for the new VM
2. Verify resource allocation matches your manifest
3. Test network connectivity if needed

Step 6: Next Steps¶

Now that you have a basic VM running, try these updates.

To make updates to deployed VMs:

1. Open the application record
2. Edit the manifest
3. Save the application
4. Reurn to the deployment record that references the application
5. Deploy and watch the updated manifest apply!

Note: the update process will typically power down the VM, make changes, and restore the VM running state to the desired setting.

Warning: Updating the name will create a new VM. Updating the VM name is currently not supported.

Add Cloud-Init Configuration¶

spec:
  resources:
    - type: virdomain
      spec:
        # ... other configuration
        cloud_init_data:
          user_data: |
            #cloud-config
            package_update: true
            packages:
              - nginx
              - htop
            runcmd:
              - systemctl enable nginx
              - systemctl start nginx
          meta_data: |
            instance-id: quickstart-vm-001
            local-hostname: quickstart-vm

Scale Your Resources¶

spec:
  resources:
    - type: virdomain
      spec:
        cpu: 2                    # Double the CPU
        memory: "4 GiB"      # Double the RAM
        storage_devices:
          - name: "storage-disk"
            type: "virtio_disk"
            capacity: "20 GB"  # Double the Storage
            boot: 1

Troubleshooting¶

Common Issues¶

1. Deployment Fails¶

Symptom: Manifest fails to deploy

Solutions: - Check YAML syntax - Verify resource availability - Ensure unique names - Check asset URLs

2. VM Won't Start¶

Symptom: VM remains in "shutoff" state

Solutions: - Verify state: "running" is set - Check resource constraints - Review boot device configuration - Check HyperCore cluster capacity

3. Network Issues¶

Symptom: VM has no network connectivity

Solutions: - Verify network device configuration - Check VirtIO driver support - Validate network settings in HyperCore

Getting Help¶

If you encounter issues:

1. Check the logs in Fleet Manager UI
2. Browse Examples for similar configurations
3. Open an issue on GitHub

What's Next?¶

Now that you've deployed your first VM, explore these advanced topics:

📚 Learn More¶

• Manifest Specification - Complete reference
• Examples - Real-world use cases
• Best Practices - Optimization tips

🚀 Try Advanced Features¶

• Multi-VM Applications - Deploy multiple VMs
• Kubernetes Deployment - Container orchestration
• GPU Applications - Machine learning workloads

🔧 Optimize Your Deployments¶

• Cloud-Init Guide - Automated configuration

Summary¶

Congratulations! You've successfully:

• ✅ Created your first Kraken manifest
• ✅ Deployed a virtual machine
• ✅ Learned basic troubleshooting
• ✅ Explored next steps

You're now ready to build more complex applications with Kraken. Happy deploying! 🎉