Skip to content

MULTI_DEPLOYMENT_AUTOSCALING.md

Latest commit

 

History

History
346 lines (250 loc) · 9.11 KB

MULTI_DEPLOYMENT_AUTOSCALING.md

File metadata and controls

346 lines (250 loc) · 9.11 KB

Multi-Deployment Support and Cloud Auto-Scaling

This control panel now supports multiple deployment methods and automatic scaling across various cloud providers, making it competitive with cPanel, Plesk, and DirectAdmin.

Deployment Methods

The control panel automatically detects and supports three deployment methods:

1. Kubernetes (Recommended for Production)

Features:

  • Full container orchestration
  • Automatic horizontal and vertical scaling
  • Load balancing
  • Self-healing deployments
  • Rolling updates with zero downtime

Supported Platforms:

  • Self-managed Kubernetes clusters
  • Amazon EKS (Elastic Kubernetes Service)
  • Microsoft Azure AKS (Azure Kubernetes Service)
  • Google GKE (Google Kubernetes Engine)
  • DigitalOcean Kubernetes (DOKS)
  • OVHcloud Managed Kubernetes

Auto-Scaling Support:

  • ✅ Horizontal Pod Autoscaling (HPA) - scales number of pods based on CPU/memory
  • ✅ Vertical Pod Autoscaling (VPA) - adjusts resource limits per pod
  • ✅ Cluster Autoscaling - adds/removes nodes automatically

2. Docker Compose

Features:

  • Container-based deployment
  • Easy local development
  • Multi-container applications
  • Volume management
  • Network isolation

Best For:

  • Development environments
  • Small to medium deployments
  • Single-server setups

Limitations:

  • No automatic scaling
  • Manual load balancing required
  • Limited high-availability options

3. Standalone (Traditional)

Features:

  • Direct server deployment
  • Traditional NGINX/Apache virtual hosts
  • Direct process management
  • Standard Linux services

Best For:

  • Legacy systems
  • Simple websites
  • Environments without container support

Limitations:

  • No automatic scaling
  • Manual configuration required
  • Limited automation

Cloud Provider Detection

The system automatically detects the cloud provider using multiple methods:

  1. Kubernetes Node Labels - Reads cloud-specific labels from K8s nodes
  2. Instance Metadata - Queries cloud metadata services
  3. Environment Variables - Checks for cloud-specific environment variables

Supported Cloud Providers:

  • Amazon Web Services (AWS)
  • Microsoft Azure
  • Google Cloud Platform (GCP)
  • DigitalOcean
  • OVHcloud
  • Linode
  • Vultr
  • Hetzner Cloud

Auto-Scaling Configuration

Global Settings

Navigate to Admin Panel → System → Deployment to view and configure:

  • Current deployment mode (automatically detected)
  • Cloud provider (automatically detected)
  • Auto-scaling status
  • System capabilities

Per-Domain Scaling

For each domain/website, you can configure:

Horizontal Pod Autoscaling (HPA)

Automatically scales the number of pods based on resource usage:

  • Minimum Replicas: Minimum number of pods (default: 1)
  • Maximum Replicas: Maximum number of pods (default: 10)
  • Target CPU: CPU utilization percentage to trigger scaling (default: 80%)

Example: If CPU usage exceeds 80%, new pods are automatically created until max replicas is reached.

Vertical Pod Autoscaling (VPA)

Automatically adjusts CPU and memory limits for pods:

  • Off: Only provides recommendations, no automatic changes
  • Initial: Applies recommendations when pods are created
  • Recreate: Deletes and recreates pods with new limits
  • Auto: Automatically updates pods (may cause brief downtime)

Manual Scaling

You can also manually set the exact number of replicas:

  • Set replicas to 0 to stop the application
  • Set replicas to 1+ to start/scale the application

Using Auto-Scaling

  1. Navigate to App Panel → Domains
  2. Select a domain
  3. Click Actions → Manage Scaling
  4. Configure your scaling preferences
  5. Click Save

The system will automatically:

  • Create HorizontalPodAutoscaler resources
  • Create VerticalPodAutoscaler resources (if enabled)
  • Monitor and scale your application

Deployment Methods Comparison

Feature Kubernetes Docker Compose Standalone
Auto-Scaling ✅ HPA + VPA ❌ ❌
Load Balancing ✅ Built-in ⚠️ Manual ⚠️ Manual
High Availability ✅ Multi-node ⚠️ Limited ❌
Zero-Downtime Deploys ✅ Rolling ⚠️ With config ❌
Resource Isolation ✅ Strong ✅ Good ⚠️ Limited
SSL Automation ✅ cert-manager ✅ Let's Encrypt ✅ Certbot
Complexity High Medium Low
Resource Overhead Medium Low Minimal

API Reference

Deployment Detection Service

use App\Services\DeploymentDetectionService;

$service = app(DeploymentDetectionService::class);

// Get deployment mode
$mode = $service->detectDeploymentMode();
// Returns: 'standalone', 'docker-compose', or 'kubernetes'

// Get cloud provider
$provider = $service->detectCloudProvider();
// Returns: 'aws', 'azure', 'gcp', 'digitalocean', etc.

// Get full deployment info
$info = $service->getDeploymentInfo();
// Returns array with mode, provider, capabilities, etc.

// Check specific environment
$isK8s = $service->isKubernetes();
$isDocker = $service->isDocker();
$isStandalone = $service->isStandalone();

// Check auto-scaling support
$canScale = $service->supportsAutoScaling();

Cloud Provider Manager

use App\Services\CloudProviderManager;

$manager = app(CloudProviderManager::class);

// Get provider for a server
$provider = $manager->getProvider($server);

// Enable horizontal scaling
$provider->enableHorizontalScaling(
    domain: $domain,
    minReplicas: 2,
    maxReplicas: 10,
    targetCpuUtilization: 75
);

// Enable vertical scaling
$provider->enableVerticalScaling(
    domain: $domain,
    options: ['update_mode' => 'Auto']
);

// Get current scaling config
$config = $provider->getScalingConfig($domain);

// Manual scaling
$provider->scaleToReplicas($domain, 5);

// Get resource metrics
$metrics = $provider->getResourceMetrics($domain);

Deployment-Aware Service

use App\Services\DeploymentAwareService;

$service = app(DeploymentAwareService::class);

// Deploy domain (automatically uses correct method)
$service->deployDomain($domain, $options);

// Delete domain
$service->deleteDomain($domain);

// Get deployment status
$status = $service->getDeploymentStatus($domain);

// Restart deployment
$service->restartDomain($domain);

Migration from Other Panels

From cPanel

The control panel provides similar functionality to cPanel:

cPanel Feature Control Panel Equivalent
WHM Admin Panel
User Panel App Panel
AutoSSL cert-manager / Let's Encrypt
Resource Limits Kubernetes Resource Quotas
Email Accounts Mail Services (Postfix/Dovecot)
FTP Accounts SFTP/FTP Services
File Manager File Manager Service
Databases MySQL/PostgreSQL Services

From Plesk

Similar to cPanel, Plesk features map to:

Plesk Feature Control Panel Equivalent
Service Plans Hosting Plans
Docker Support Native Docker/K8s
Git Deployment Git Deployment Service
WordPress Toolkit WordPress Auto-Deployment
Security NetworkPolicies + RBAC

From DirectAdmin

DirectAdmin users will find:

DirectAdmin Feature Control Panel Equivalent
User Levels Team-based RBAC
Reseller Accounts Multi-tenancy Support
Backup/Restore Backup Services
Custom NS DNS Cluster

Best Practices

Production Deployments

  1. Use Kubernetes for production workloads
  2. Enable Auto-Scaling for traffic spikes
  3. Set Resource Limits to prevent resource exhaustion
  4. Use VPA with "Initial" mode to avoid disruption
  5. Monitor Metrics regularly

Development Environments

  1. Use Docker Compose for local development
  2. Match production config as closely as possible
  3. Use volume mounts for faster iterations

Resource Limits

Small Sites (< 1000 visitors/day)

  • Min Replicas: 1
  • Max Replicas: 3
  • CPU Target: 80%
  • Memory: 256Mi - 512Mi

Medium Sites (1000-10000 visitors/day)

  • Min Replicas: 2
  • Max Replicas: 5
  • CPU Target: 70%
  • Memory: 512Mi - 1Gi

Large Sites (> 10000 visitors/day)

  • Min Replicas: 3
  • Max Replicas: 10+
  • CPU Target: 60%
  • Memory: 1Gi - 2Gi

Troubleshooting

Auto-Scaling Not Working

  1. Check deployment mode: Admin Panel → Deployment
  2. Verify Kubernetes is detected
  3. Verify cloud provider is detected
  4. Check metrics-server is installed: kubectl get deployment metrics-server -n kube-system

Pods Not Scaling

  1. Check HPA status: Navigate to domain → Manage Scaling
  2. Verify resource requests are set
  3. Check pod metrics: View resource metrics in scaling dialog
  4. Ensure CPU/memory thresholds are reached

Deployment Fails

  1. Check server type matches deployment method
  2. Verify server credentials
  3. Check logs in deployment status
  4. Ensure required services are running

Support

For issues or questions:

  1. Check GitHub Issues: https://github.com/liberu-control-panel/control-panel-laravel/issues
  2. Read documentation: https://liberu.co.uk
  3. Community support via GitHub Discussions

License

This feature is part of the Liberu Control Panel and is licensed under the MIT License.