Upgrading

This guide covers upgrading LXC AutoScale ML from previous versions.

Upgrade from 1.x to 2.0

Version 2.0 introduces significant improvements including batch async API, enhanced security, and critical bug fixes.

Step 1: Backup Configuration

cp /etc/lxc_autoscale_ml/lxc_autoscale_ml.yaml /etc/lxc_autoscale_ml/lxc_autoscale_ml.yaml.backup
cp /etc/lxc_autoscale_ml/lxc_autoscale_api.yaml /etc/lxc_autoscale_ml/lxc_autoscale_api.yaml.backup
cp /etc/lxc_autoscale_ml/lxc_monitor.yaml /etc/lxc_autoscale_ml/lxc_monitor.yaml.backup

Step 2: Stop Services

systemctl stop lxc_autoscale_api lxc_monitor lxc_autoscale_ml

Step 3: Install New Dependencies

apt install -y python3-aiohttp
# Optional: apt install -y python3-prometheus-client

Step 4: Update Application Files

cd /tmp
git clone https://github.com/fabriziosalmi/proxmox-lxc-autoscale-ml.git
cp -r lxc-autoscale-ml/lxc_autoscale_ml/* /usr/local/bin/lxc_autoscale_ml/

Step 5: Update Configuration

Add new settings to /etc/lxc_autoscale_ml/lxc_autoscale_api.yaml:

authentication:
  enabled: true
  api_keys:
    - "your-secret-key-here"  # Generate with: openssl rand -hex 32

rate_limiting:
  enabled: true
  max_requests_per_minute: 120  # Increased from 60
  time_window_seconds: 60

Add new settings to /etc/lxc_autoscale_ml/lxc_autoscale_ml.yaml:

scaling:
  cpu_scale_step: 1           # NEW: Incremental scaling
  ram_scale_step_mb: 512      # NEW: Incremental scaling

ignore_lxc: []                # CHANGED: Was ["101", "102"]

circuit_breaker:              # NEW section
  enabled: true
  failure_threshold: 5
  timeout_seconds: 300

Add new settings to /etc/lxc_autoscale_ml/lxc_monitor.yaml:

monitoring:
  max_metrics_entries: 1000   # NEW: bounds the metrics file

Step 6: Start Services

systemctl start lxc_autoscale_api lxc_monitor lxc_autoscale_ml

Step 7: Verify Upgrade

# Check services are running
systemctl status lxc_autoscale_api lxc_autoscale_ml lxc_monitor

# Test API with authentication
curl -H "X-API-Key: your-secret-key-here" http://127.0.0.1:5000/health/check

# Monitor logs for batch fetch
journalctl -u lxc_autoscale_ml -f | grep "Batch fetch"

Breaking Changes in 2.0

API Authentication Required

All endpoints except /health/check and /metrics now require authentication.

Before (1.x):

curl http://proxmox:5000/resource/lxc/status?lxc_id=104

After (2.0):

curl -H "X-API-Key: your-key" http://proxmox:5000/resource/lxc/status?lxc_id=104

Default Ignore List Changed

The default ignore_lxc list is now empty. Previously, containers 101 and 102 were ignored by default.

Action required: Review your ignore_lxc setting and add any containers you want to exclude from scaling.

Rate Limit Increased

The default rate limit increased from 60 to 120 requests per minute. Localhost requests are now exempt from rate limiting.

Update Scripts and Automation

If you have scripts calling the API, update them to include authentication:

# Old
curl -X POST http://proxmox:5000/scale/cores \
  -H "Content-Type: application/json" \
  -d '{"lxc_id": 104, "cores": 4}'

# New
curl -X POST http://proxmox:5000/scale/cores \
  -H "Content-Type: application/json" \
  -H "X-API-Key: your-key" \
  -d '{"lxc_id": 104, "cores": 4}'

Rollback Procedure

If you need to rollback to version 1.x:

# Stop services
systemctl stop lxc_autoscale_api lxc_monitor lxc_autoscale_ml

# Restore configuration
cp /etc/lxc_autoscale_ml/lxc_autoscale_ml.yaml.backup /etc/lxc_autoscale_ml/lxc_autoscale_ml.yaml
cp /etc/lxc_autoscale_ml/lxc_autoscale_api.yaml.backup /etc/lxc_autoscale_ml/lxc_autoscale_api.yaml

# Restore old application files from your backup or re-download 1.x release

# Start services
systemctl start lxc_autoscale_api lxc_monitor lxc_autoscale_ml

Verify Performance Improvements

After upgrading, you should see significant performance improvements in the logs:

# Before (1.x) - Sequential fetching
INFO - Fetching config for container 101...
INFO - Fetching config for container 102...
# Takes ~6 seconds for 60 containers

# After (2.0) - Batch async fetching
INFO - Batch fetching configs for 60 containers...
INFO - Batch fetch completed in 0.58s: 60/60 successful (103.4 containers/sec)

Next Steps

• Configuration: Review all new configuration options