Configuration Reference

This page documents all configuration options for LXC AutoScale ML.

Configuration Files

Component	Path
API	/etc/lxc_autoscale_ml/lxc_autoscale_api.yaml
Model	/etc/lxc_autoscale_ml/lxc_autoscale_ml.yaml
Monitor	/etc/lxc_autoscale_ml/lxc_monitor.yaml

API Configuration

Complete Reference

# /etc/lxc_autoscale_ml/lxc_autoscale_api.yaml

# Server Configuration
server:
  host: "0.0.0.0"          # Bind address
  port: 5000               # Listen port

# Authentication
authentication:
  enabled: true            # Enable API key authentication
  api_key: "your-key"      # API key (change this!)

# Rate Limiting
rate_limiting:
  enabled: true            # Enable rate limiting
  max_requests_per_minute: 120  # Max requests per IP
  time_window_seconds: 60  # Sliding window duration

# Logging
logging:
  log_level: "INFO"        # DEBUG, INFO, WARNING, ERROR
  log_file: "/var/log/autoscaleapi.log"
  access_log: "/var/log/autoscaleapi_access.log"
  error_log: "/var/log/autoscaleapi_error.log"

API Options

Option	Type	Default	Description
server.host	string	0.0.0.0	Network interface to bind
server.port	integer	5000	TCP port number
authentication.enabled	boolean	true	Require API key
authentication.api_key	string	-	Secret API key
rate_limiting.enabled	boolean	true	Enable rate limits
rate_limiting.max_requests_per_minute	integer	120	Request limit per IP
rate_limiting.time_window_seconds	integer	60	Window duration
logging.log_level	string	INFO	Minimum log level

Model Configuration

Complete Reference

# /etc/lxc_autoscale_ml/lxc_autoscale_ml.yaml

# API Configuration
api:
  api_url: "http://127.0.0.1:5000"  # API base URL
  timeout: 5               # Request timeout (seconds)
  max_concurrent: 10       # Max parallel requests
  retry_attempts: 3        # Retry failed requests

# Data Configuration
data:
  metrics_file: "/var/log/lxc_metrics.json"

# Logging Configuration
logging:
  log_level: "INFO"        # DEBUG, INFO, WARNING, ERROR
  log_file: "/var/log/lxc_autoscale_ml.log"

# IsolationForest Configuration
isolation_forest:
  contamination: 0.1       # Expected anomaly fraction
  n_estimators: 100        # Number of trees
  random_state: 42         # Random seed
  max_samples: "auto"      # Samples per tree

# Scaling Configuration
scaling:
  # CPU Thresholds (percentage)
  cpu_scale_up_threshold: 70
  cpu_scale_down_threshold: 30
  cpu_scale_step: 1        # Cores to add/remove

  # RAM Thresholds (percentage)
  ram_scale_up_threshold: 80
  ram_scale_down_threshold: 40
  ram_scale_step_mb: 512   # MB to add/remove

  # Resource Limits
  min_cpu_cores: 1
  max_cpu_cores: 8
  min_ram_mb: 512
  max_ram_mb: 16384

  # Confidence
  min_confidence: 70       # Minimum confidence to scale

  # Testing
  dry_run: false           # Log decisions without executing

# Ignored Containers
ignore_lxc: []             # List of container IDs to skip

# Circuit Breaker
circuit_breaker:
  enabled: true
  failure_threshold: 5     # Failures before opening
  timeout_seconds: 300     # Time before reset

# Sleep Configuration
sleep_interval: 60         # Seconds between cycles

Model Options

API Settings

Option	Type	Default	Description
api.api_url	string	http://127.0.0.1:5000	API base URL
api.timeout	integer	5	Request timeout in seconds
api.max_concurrent	integer	10	Maximum parallel requests
api.retry_attempts	integer	3	Retry count for failures

IsolationForest Settings

Option	Type	Default	Description
isolation_forest.contamination	float	0.1	Expected anomaly ratio (0.0-0.5)
isolation_forest.n_estimators	integer	100	Number of trees
isolation_forest.random_state	integer	42	Random seed
isolation_forest.max_samples	string/int	auto	Samples per tree

Scaling Settings

Option	Type	Default	Description
scaling.cpu_scale_up_threshold	integer	70	CPU % to trigger scale up
scaling.cpu_scale_down_threshold	integer	30	CPU % to trigger scale down
scaling.cpu_scale_step	integer	1	CPU cores per scaling action
scaling.ram_scale_up_threshold	integer	80	RAM % to trigger scale up
scaling.ram_scale_down_threshold	integer	40	RAM % to trigger scale down
scaling.ram_scale_step_mb	integer	512	RAM MB per scaling action
scaling.min_cpu_cores	integer	1	Minimum CPU cores
scaling.max_cpu_cores	integer	8	Maximum CPU cores
scaling.min_ram_mb	integer	512	Minimum RAM (MB)
scaling.max_ram_mb	integer	16384	Maximum RAM (MB)
scaling.min_confidence	integer	70	Minimum scaling confidence
scaling.dry_run	boolean	false	Test mode (no actual scaling)

Circuit Breaker Settings

Option	Type	Default	Description
circuit_breaker.enabled	boolean	true	Enable circuit breaker
circuit_breaker.failure_threshold	integer	5	Failures before opening
circuit_breaker.timeout_seconds	integer	300	Reset timeout

Other Settings

Option	Type	Default	Description
ignore_lxc	list	[]	Container IDs to exclude
sleep_interval	integer	60	Seconds between cycles

Monitor Configuration

Complete Reference

# /etc/lxc_autoscale_ml/lxc_monitor.yaml

# Metrics Configuration
metrics:
  output_file: "/var/log/lxc_metrics.json"
  max_entries: 1000        # Max entries in file
  collection_interval: 10  # Seconds between collections

# Logging Configuration
logging:
  log_level: "INFO"        # DEBUG, INFO, WARNING, ERROR
  log_file: "/var/log/lxc_monitor.log"

# Container Filter
containers:
  ignore_stopped: true     # Skip stopped containers
  ignore_templates: true   # Skip template containers

# Performance
performance:
  batch_size: 10           # Containers per batch
  timeout: 5               # Timeout per container

Monitor Options

Option	Type	Default	Description
metrics.output_file	string	/var/log/lxc_metrics.json	Metrics output path
metrics.max_entries	integer	1000	Maximum entries to keep
metrics.collection_interval	integer	10	Collection interval (seconds)
containers.ignore_stopped	boolean	true	Skip stopped containers
containers.ignore_templates	boolean	true	Skip template containers
performance.batch_size	integer	10	Containers per batch
performance.timeout	integer	5	Timeout per container

Configuration Validation

The model validates configuration on startup. Common validation rules:

Rule	Error Message
min_cpu_cores > max_cpu_cores	"min_cpu_cores cannot be greater than max_cpu_cores"
min_ram_mb > max_ram_mb	"min_ram_mb cannot be greater than max_ram_mb"
cpu_scale_down_threshold >= cpu_scale_up_threshold	"cpu_scale_down_threshold must be less than cpu_scale_up_threshold"
Threshold outside 0-100	"threshold must be between 0 and 100"

Validate Configuration Syntax

python3 -c "import yaml; yaml.safe_load(open('/etc/lxc_autoscale_ml/lxc_autoscale_ml.yaml'))"

If there's no output, the YAML is valid.

Example Configurations

Conservative Scaling

Slower, more cautious scaling for production:

scaling:
  cpu_scale_up_threshold: 85
  cpu_scale_down_threshold: 20
  ram_scale_up_threshold: 90
  ram_scale_down_threshold: 30
  min_confidence: 80

sleep_interval: 120

Aggressive Scaling

Faster response for development environments:

scaling:
  cpu_scale_up_threshold: 60
  cpu_scale_down_threshold: 40
  ram_scale_up_threshold: 70
  ram_scale_down_threshold: 50
  min_confidence: 60

sleep_interval: 30

Large Deployment

Optimized for 60+ containers:

api:
  max_concurrent: 15
  timeout: 10

circuit_breaker:
  failure_threshold: 3
  timeout_seconds: 600

sleep_interval: 120

Applying Configuration Changes

After modifying configuration files, restart the relevant service:

# API changes
systemctl restart lxc_autoscale_api

# Model changes
systemctl restart lxc_autoscale_ml

# Monitor changes
systemctl restart lxc_monitor