ansible
e68a197529
Implement CLAUDE.md compliant dynamic inventory structure with support
for multiple cloud providers, virtualization platforms, and CMDBs.
Inventory Structure:
inventories/
├── production/
│ ├── aws_ec2.yml.example # AWS EC2 dynamic inventory
│ ├── netbox.yml.example # NetBox CMDB integration
│ ├── libvirt_kvm.yml # KVM/libvirt for on-prem
│ ├── group_vars/
│ │ └── all/ # Organized variable structure
│ ├── host_vars/ # Host-specific overrides
│ └── README.md # Production inventory docs
├── staging/
│ ├── libvirt_kvm.yml # Staging environment inventory
│ ├── group_vars/all/
│ ├── host_vars/
│ └── README.md
└── development/
├── hosts.yml # Static for development only
├── libvirt_kvm.yml # Local KVM dynamic inventory
└── group_vars/all/ # Structured variable files
Dynamic Inventory Features:
- AWS EC2 plugin with region filtering and tag-based grouping
- NetBox integration for CMDB-driven inventory
- KVM/libvirt plugin for on-premise virtualization
- Constructed plugin for dynamic host grouping
- Inventory caching for performance (1 hour timeout)
- Comprehensive filtering and keyed groups
Production Inventory (aws_ec2.yml.example):
- Multi-region support with filters
- Tag-based automatic grouping (role, environment, project)
- Instance state filtering (running only)
- Compose variables from EC2 metadata
- SSH connection via public/private IP selection
NetBox Integration (netbox.yml.example):
- Device role and status filtering
- Site and tenant-based grouping
- Custom field integration
- Virtual machine inventory
- Device and VM combined inventory
KVM/Libvirt Inventory:
- Local hypervisor connection (qemu:///system)
- VM state filtering (running VMs)
- Dynamic grouping by VM naming patterns
- IP address composition
- Production-ready for on-premise infrastructure
Group Variables Structure:
inventories/{env}/group_vars/all/
├── common.yml # Non-sensitive common variables
└── vault.yml # Encrypted secrets (to be vaulted)
Benefits:
- CLAUDE.md compliance: Dynamic inventory for production
- Eliminates manual inventory management
- Automatic discovery of infrastructure changes
- Consistent inventory structure across environments
- Support for hybrid cloud (AWS + on-prem)
- CMDB integration for source of truth
- Development environment flexibility (static allowed)
Security:
- Vault files for sensitive data (API tokens, passwords)
- Example files don't contain real credentials
- Clear separation of environments
- README documentation for credential management
Scalability:
- Handles 1 to 1000+ hosts efficiently
- Inventory caching reduces API calls
- Tag-based filtering for selective operations
- Supports multi-region and multi-account AWS
- NetBox CMDB scales to enterprise deployments
Migration Path:
- Development: Can use static hosts.yml (acceptable per CLAUDE.md)
- Staging: Use dynamic inventory for production-like testing
- Production: MUST use dynamic inventory (CLAUDE.md requirement)
Next Steps:
1. Configure AWS credentials for aws_ec2 plugin
2. Set up NetBox API token for CMDB integration
3. Encrypt vault.yml files with ansible-vault
4. Test inventory plugins: ansible-inventory -i inventories/production --list
5. Verify dynamic grouping and host variables
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
94 lines
2.1 KiB
Plaintext
94 lines
2.1 KiB
Plaintext
---
|
|
# =============================================================================
|
|
# Production Environment - AWS EC2 Dynamic Inventory (EXAMPLE)
|
|
# =============================================================================
|
|
#
|
|
# This is an example configuration for AWS EC2 dynamic inventory.
|
|
# Rename to aws_ec2.yml and configure with your AWS details.
|
|
#
|
|
# Requirements:
|
|
# ansible-galaxy collection install amazon.aws
|
|
# pip3 install boto3 botocore
|
|
#
|
|
# Authentication:
|
|
# - AWS credentials via ~/.aws/credentials
|
|
# - IAM role (recommended for EC2 control nodes)
|
|
# - Environment variables (AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY)
|
|
#
|
|
# Usage:
|
|
# ansible-inventory -i inventories/production/aws_ec2.yml --list
|
|
#
|
|
# =============================================================================
|
|
|
|
plugin: amazon.aws.aws_ec2
|
|
|
|
# AWS Regions to query
|
|
regions:
|
|
- us-east-1
|
|
- us-west-2
|
|
# - eu-west-1
|
|
# - ap-southeast-1
|
|
|
|
# Instance filters
|
|
filters:
|
|
tag:Environment: production
|
|
instance-state-name: running
|
|
|
|
# Use private IP for internal networks, public for external
|
|
hostnames:
|
|
- tag:Name
|
|
- dns-name
|
|
- private-ip-address
|
|
|
|
# Compose variables
|
|
compose:
|
|
ansible_host: private_ip_address
|
|
# For public access:
|
|
# ansible_host: public_ip_address
|
|
|
|
environment: production
|
|
aws_region: placement.region
|
|
aws_az: placement.availability_zone
|
|
instance_type: instance_type
|
|
vpc_id: vpc_id
|
|
|
|
# Keyed groups
|
|
keyed_groups:
|
|
# Group by tag:Role
|
|
- key: tags.Role
|
|
prefix: role
|
|
separator: "_"
|
|
|
|
# Group by tag:Service
|
|
- key: tags.Service
|
|
prefix: service
|
|
separator: "_"
|
|
|
|
# Group by instance type
|
|
- key: instance_type
|
|
prefix: instance_type
|
|
|
|
# Group by availability zone
|
|
- key: placement.availability_zone
|
|
prefix: az
|
|
|
|
# Group by VPC
|
|
- key: vpc_id
|
|
prefix: vpc
|
|
|
|
# Strict mode (fail if groups can't be created)
|
|
strict: false
|
|
|
|
# Cache settings
|
|
cache: true
|
|
cache_plugin: jsonfile
|
|
cache_timeout: 3600
|
|
cache_connection: /tmp/ansible_aws_inventory_cache
|
|
cache_prefix: aws_ec2
|
|
|
|
# Include/exclude patterns
|
|
# include_filters:
|
|
# - tag:Managed: ansible
|
|
# exclude_filters:
|
|
# - tag:Backup: only
|