lm-cloud-inventory 2.0.0

Cloud resource inventory collection for LogicMonitor licensing

LM Cloud Resource Inventory

A unified solution for collecting cloud resource counts across AWS, Azure, GCP, and OCI for LogicMonitor licensing purposes.

Overview

This tool collects resource inventory from cloud providers and calculates LogicMonitor license requirements by categorizing resources into:

• IaaS - Virtual machines and compute instances
• PaaS - Managed services, containers, serverless functions
• Non-Compute - Storage, networking, and other infrastructure resources

Quick Start

Installation

pip install lm-cloud-inventory

Or install from source:

git clone https://github.com/logicmonitor/lm-cloud-resource-inventory.git
cd lm-cloud-resource-inventory
pip install .

Basic Usage

# Run inventory and calculate licenses
lmci run -p aws -o aws_summary.csv
lmci run -p azure -o azure_summary.csv
lmci run -p gcp -o gcp_summary.csv
lmci run -p oci -o oci_summary.csv

---

CLI Reference

Global Options

Option	Description
-v, --verbose	Enable verbose logging (shows debug info and full tracebacks)
--version	Show version and exit
--help	Show help message

run Command (Recommended)

Collect inventory and calculate licenses in one step. This is the primary command for most users.

lmci run -p <provider> [options]

Option	Provider	Description
-p, --provider	All	Cloud provider: aws, azure, gcp, oci (required)
-o, --output	All	Output CSV file (default: license_summary.csv)
-d, --detailed	All	Generate detailed CSV with per-region breakdown
--show-unmapped	All	List resource types not mapped to license categories
--profile	AWS	AWS CLI profile name
-s, --subscription	Azure	Subscription ID (can be repeated for multiple)
--project	GCP	GCP project ID
--compartment	OCI	OCI compartment OCID

Examples:

# AWS - basic
lmci run -p aws

# AWS - with named profile
lmci run -p aws --profile production -o aws_prod.csv

# Azure - all subscriptions
lmci run -p azure

# Azure - specific subscriptions
lmci run -p azure -s "sub-id-1" -s "sub-id-2"

# GCP - auto-detect project from credentials
lmci run -p gcp

# GCP - explicit project
lmci run -p gcp --project my-project-id

# OCI - entire tenancy
lmci run -p oci

# OCI - specific compartment
lmci run -p oci --compartment ocid1.compartment.oc1..xxx

# Any provider - with detailed output
lmci run -p aws -d -o detailed_report.csv

collect Command

Collect inventory only (saves JSON for later calculation).

lmci collect -p <provider> [options]

Option	Provider	Description
-p, --provider	All	Cloud provider (required)
-o, --output	All	Output JSON file (default: inventory.json)
--profile	AWS	AWS CLI profile name
--region	AWS	AWS region for API calls (default: us-east-1)
--organization	AWS	IAM role name for cross-account access
--organization	GCP	GCP organization ID for org-wide inventory
-s, --subscription	Azure	Subscription ID (repeatable)
--project	GCP	GCP project ID
--compartment	OCI	OCI compartment OCID

calculate Command

Calculate licenses from an existing inventory JSON file.

lmci calculate -i <inventory.json> [options]

Option	Description
-i, --input	Input inventory JSON file (required)
-o, --output	Output CSV file (default: license_summary.csv)
-d, --detailed	Generate detailed CSV
--show-unmapped	List unmapped resource types

permissions Command

Show required permissions for each cloud provider.

lmci permissions -p aws
lmci permissions -p azure
lmci permissions -p gcp
lmci permissions -p oci

---

Supported Cloud Providers

Provider	API Used	Performance
AWS	AWS Resource Explorer	~2-5 minutes
Azure	Azure Resource Graph	~1-2 minutes
GCP	Cloud Asset Inventory	~1-2 minutes
OCI	OCI Search Service	~1 minute

For a complete list of supported resources, see LogicMonitor Cloud Services Documentation.

AWS Resource Explorer Limitations

The following AWS services are not indexed by AWS Resource Explorer and will not be collected:

Service	Resource Type
CloudSearch	Domain
MediaConnect	Flow
MediaConvert	Queue
OpsWorks	Stack
Q Business	Application
QuickSight	Dashboard (dataset/datasource are supported)
Simple Workflow (SWF)	Domain
Application Migration Service	Source Server
ElasticTranscoder	Pipeline

If you have significant usage of these services, contact your LogicMonitor representative for manual inventory assistance.

---

Requirements

• Python 3.9+
• Cloud provider credentials configured
• Read-only permissions (see Permissions Documentation)

---

Credential Setup

Configure credentials for each cloud provider before running the inventory.

AWS Credentials

Option 1: AWS CLI (Recommended)

# Install AWS CLI and configure
aws configure

# Verify
aws sts get-caller-identity

Option 2: Environment Variables

export AWS_ACCESS_KEY_ID="your-access-key"
export AWS_SECRET_ACCESS_KEY="your-secret-key"
export AWS_DEFAULT_REGION="us-east-1"

Option 3: Named Profiles

# Use a specific profile
lmci run -p aws --profile production

Resources: AWS CLI Configuration

---

Azure Credentials

Option 1: Azure CLI (Recommended)

# Sign in
az login

# Verify
az account list --output table

Option 2: Service Principal

export AZURE_CLIENT_ID="appId"
export AZURE_CLIENT_SECRET="password"
export AZURE_TENANT_ID="tenant"

Resources: Azure CLI Authentication

---

GCP Credentials

Option 1: gcloud CLI (Recommended)

# Initialize and authenticate
gcloud init
gcloud auth application-default login

# Verify
gcloud config list

Option 2: Service Account Key

export GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account-key.json"

The tool will automatically extract the project_id from the service account JSON file.

Resources: GCP Application Default Credentials

---

OCI Credentials

Option 1: OCI CLI (Recommended)

# Run setup wizard
oci setup config

# Verify
oci iam region list

Configuration is stored in ~/.oci/config.

Resources: OCI CLI Quickstart

---

Running in Cloud Shell

Each cloud provider offers a browser-based shell with credentials pre-configured - the fastest way to run the inventory tool.

Provider	Cloud Shell	Notes
AWS	AWS CloudShell	Python pre-installed, credentials automatic
Azure	Azure Cloud Shell	Python pre-installed, az authenticated
GCP	Google Cloud Shell	Python & gcloud pre-installed
OCI	OCI Cloud Shell	OCI CLI pre-installed

---

Output Files

Summary CSV

The default output includes resource type breakdown by category:

Provider,Account,Category,ResourceType,Region,Count
aws,123456789012,IaaS,ec2:instance,us-east-1,42
aws,123456789012,IaaS,ec2:instance,us-west-2,15
aws,123456789012,PaaS,lambda:function,us-east-1,28

TOTAL,,IaaS,,,57
TOTAL,,PaaS,,,28
TOTAL,,Non-Compute,,,125

Detailed CSV (with -d flag)

Same as summary but saved to a separate _detailed.csv file.

Raw Inventory JSON

The _inventory.json file contains raw resource data for analysis:

[
  {
    "provider": "aws",
    "account_id": "123456789012",
    "region": "us-east-1",
    "resource_type": "ec2:instance",
    "count": 42,
    "timestamp": "2024-12-22T10:30:00Z"
  }
]

---

Troubleshooting

AWS: "Resource Explorer not enabled"

AWS Resource Explorer is required. Enable it in your account:

1. Go to AWS Resource Explorer Console
2. Enable Resource Explorer
3. Create an aggregator index for cross-region queries

AWS: "Access Denied"

# Verify credentials
aws sts get-caller-identity

Check Permissions Documentation for required IAM permissions.

Azure: "AuthorizationFailed"

# Verify login
az account show

Ensure you have the Reader role on the subscription(s).

GCP: "Permission denied"

# Verify auth
gcloud auth list

Ensure you have the roles/cloudasset.viewer role.

OCI: "NotAuthorizedOrNotFound"

# Verify config
cat ~/.oci/config

Ensure the policy allows: Allow group <group> to inspect all-resources in tenancy

General: "ModuleNotFoundError"

pip install .

---

Security & Privacy

Security:

• Read-only access only - no write/modify/delete operations
• Uses provider credential chains - no credential storage
• Only communicates with cloud provider APIs

Data Collected:

• Resource type counts
• Account/subscription identifiers
• Region/location information

NOT Collected:

• Resource names or IDs
• Resource content or data
• Configuration details
• Credentials or secrets

---

Documentation

• Required Permissions
• LogicMonitor Cloud Services

---

Support

• Pre-sales: Contact your LogicMonitor Sales Engineer
• Customers: Contact your Customer Success Manager

---

License

MIT License - See LICENSE file for details.