Configuration Options - antimetal/system-agent GitHub Wiki

Configuration Options

⚠️ Work in Progress: This documentation is currently being developed and may be incomplete or subject to change.

Overview

This page provides a comprehensive reference of all configuration options available for the Antimetal System Agent. Configuration can be provided through command-line flags, environment variables, or configuration files.

Configuration Methods

Priority Order

Configuration is applied in the following order (highest priority first):

Command-line flags
Environment variables
Configuration file
Default values

Command-Line Flags

antimetal-agent \
  --cluster-name=production \
  --api-endpoint=grpc.antimetal.com:443 \
  --collection-interval=30s \
  --log-level=info

Environment Variables

All flags can be set via environment variables with the prefix ANTIMETAL_:

export ANTIMETAL_CLUSTER_NAME=production
export ANTIMETAL_API_ENDPOINT=grpc.antimetal.com:443
export ANTIMETAL_COLLECTION_INTERVAL=30s
export ANTIMETAL_LOG_LEVEL=info

Configuration File

YAML configuration file specified with --config flag:

cluster_name: production
api_endpoint: grpc.antimetal.com:443
collection_interval: 30s
log_level: info

Core Configuration

Basic Settings

Option	Type	Default	Description
`cluster_name`	string	(required)	Kubernetes cluster identifier
`api_endpoint`	string	`grpc.antimetal.com:443`	Antimetal API endpoint
`api_key`	string	(required)	API authentication key
`disable_tls`	bool	`false`	Disable TLS (development only)

Logging Configuration

Option	Type	Default	Description
`log_level`	string	`info`	Log level (debug, info, warn, error)
`log_format`	string	`json`	Log format (json, text)
`log_output`	string	`stdout`	Log output (stdout, stderr, file)
`log_file`	string	-	Log file path (when output=file)

Collection Configuration

Kubernetes Resources

Option	Type	Default	Description
`kube_config`	string	In-cluster	Path to kubeconfig file
`kube_context`	string	Current	Kubernetes context to use
`namespaces`	[]string	All	Namespaces to monitor
`exclude_namespaces`	[]string	-	Namespaces to exclude
`resource_types`	[]string	All	Resource types to collect

Performance Metrics

Option	Type	Default	Description
`enable_performance`	bool	`true`	Enable performance collectors
`collection_interval`	duration	`10s`	Metric collection interval
`host_proc_path`	string	`/proc`	Path to proc filesystem
`host_sys_path`	string	`/sys`	Path to sys filesystem

Collector-Specific Options

CPU Collector

Option	Type	Default	Description
`cpu.enabled`	bool	`true`	Enable CPU metrics
`cpu.interval`	duration	`10s`	CPU collection interval
`cpu.per_core`	bool	`true`	Collect per-core stats

Memory Collector

Option	Type	Default	Description
`memory.enabled`	bool	`true`	Enable memory metrics
`memory.interval`	duration	`10s`	Memory collection interval
`memory.include_swap`	bool	`true`	Include swap metrics

Disk Collector

Option	Type	Default	Description
`disk.enabled`	bool	`true`	Enable disk metrics
`disk.interval`	duration	`30s`	Disk collection interval
`disk.exclude_devices`	[]string	-	Devices to exclude

Network Collector

Option	Type	Default	Description
`network.enabled`	bool	`true`	Enable network metrics
`network.interval`	duration	`10s`	Network collection interval
`network.exclude_interfaces`	[]string	`["lo"]`	Interfaces to exclude

Process Collector

Option	Type	Default	Description
`process.enabled`	bool	`true`	Enable process metrics
`process.interval`	duration	`30s`	Process collection interval
`process.top_count`	int	`20`	Number of top processes

Advanced Configuration

Performance Tuning

Option	Type	Default	Description
`worker_threads`	int	`4`	Number of worker threads
`batch_size`	int	`1000`	Resource batch size
`queue_size`	int	`10000`	Internal queue size
`send_timeout`	duration	`30s`	API send timeout

Filtering Options

filters:
  # Label-based filtering
  labels:
    include:
      environment: ["production", "staging"]
      monitored: ["true"]
    exclude:
      temporary: ["true"]
  
  # Annotation-based filtering
  annotations:
    exclude:
      - "kubectl.kubernetes.io/last-applied-configuration"
  
  # Resource name patterns
  names:
    exclude:
      - "^test-.*"
      - ".*-debug$"

Cloud Provider Settings

AWS/EKS

Option	Type	Default	Description
`cloud.provider`	string	Auto-detect	Cloud provider (aws, gcp, azure)
`aws.region`	string	Auto-detect	AWS region
`aws.assume_role`	string	-	IAM role to assume

GCP/GKE

Option	Type	Default	Description
`gcp.project`	string	Auto-detect	GCP project ID
`gcp.zone`	string	Auto-detect	GCP zone

Azure/AKS

Option	Type	Default	Description
`azure.subscription_id`	string	Auto-detect	Azure subscription
`azure.resource_group`	string	Auto-detect	Resource group

Security Options

Option	Type	Default	Description
`tls.ca_cert`	string	System CA	Custom CA certificate
`tls.client_cert`	string	-	Client certificate
`tls.client_key`	string	-	Client key
`tls.server_name`	string	-	TLS server name override
`tls.insecure_skip_verify`	bool	`false`	Skip TLS verification

Configuration Examples

Minimal Configuration

cluster_name: my-cluster
api_key: ${ANTIMETAL_API_KEY}

Production Configuration

cluster_name: production-eks-us-west-2
api_endpoint: grpc.antimetal.com:443
api_key: ${ANTIMETAL_API_KEY}

# Logging
log_level: info
log_format: json

# Collection
collection_interval: 30s
namespaces:
  - production
  - staging

# Performance
cpu:
  interval: 10s
memory:
  interval: 10s
disk:
  interval: 60s
  exclude_devices:
    - "^loop.*"
    - "^dm-.*"

# Filtering
filters:
  labels:
    include:
      monitored: ["true"]
  namespaces:
    exclude:
      - kube-system
      - kube-public

# Performance tuning
worker_threads: 8
batch_size: 2000

Development Configuration

cluster_name: local-development
api_endpoint: localhost:8443
disable_tls: true
api_key: dev-key

log_level: debug
log_format: text

# Reduced collection for development
collection_interval: 60s
process:
  enabled: false

Environment-Specific Settings

Container Environments

When running in containers, adjust paths:

host_proc_path: /host/proc
host_sys_path: /host/sys

Kubernetes Deployment

For in-cluster deployment:

# Automatically uses in-cluster config
kube_config: ""  # Empty means in-cluster

# Service account must have appropriate RBAC

Configuration Validation

The agent validates configuration on startup:

Required fields are present
Data types are correct
Paths exist and are accessible
API connectivity is verified

Invalid configuration causes the agent to exit with an error.

Dynamic Configuration

Some settings can be changed at runtime:

Log level (via signals)
Collection intervals (via API)
Filter rules (via API)