asset-inventory – skill DevOps & Hạ tầng cho agent AI

Asset Inventory

Maintain comprehensive IT asset inventory using automated discovery, AWS Config rules, cloud asset discovery scripts, CMDB integration, and tagging enforcement for compliance and operational visibility.

Asset Categories and Schema

asset_categories:
  compute:
    cloud:
      - EC2 instances / Azure VMs / GCE instances
      - Lambda functions / Azure Functions / Cloud Functions
      - ECS/EKS clusters and tasks
      - Container images in registries
    on_premise:
      - Physical servers
      - Virtual machines (VMware, Hyper-V)

  storage:
    - S3 buckets / Azure Storage / GCS buckets
    - EBS volumes / Managed Disks / Persistent Disks
    - RDS instances / Azure SQL / Cloud SQL
    - DynamoDB tables / Cosmos DB / Firestore
    - EFS / Azure Files / Filestore

  network:
    - VPCs / VNets / VPC Networks
    - Load balancers (ALB, NLB, Azure LB, GCP LB)
    - DNS zones and records
    - VPN gateways and connections
    - CDN distributions

  security:
    - IAM users, roles, and policies
    - KMS keys / Key Vault keys
    - Certificates (ACM, Key Vault, Certificate Manager)
    - Security groups / NSGs / Firewall rules
    - WAF configurations

  applications:
    - SaaS subscriptions
    - Licensed software
    - Custom applications
    - APIs and integrations

  endpoints:
    - Laptops and desktops
    - Mobile devices
    - Printers and peripherals

asset_record_schema:
  required_fields:
    asset_id: "Unique identifier (auto-generated)"
    name: "Human-readable name"
    type: "Category from above taxonomy"
    provider: "AWS / Azure / GCP / On-Premise / SaaS"
    account_or_subscription: "Cloud account ID"
    region: "Deployment region/location"
    owner: "Team or individual responsible"
    data_classification: "Public / Internal / Confidential / Restricted"
    environment: "Production / Staging / Development / Sandbox"
    status: "Active / Decommissioning / Retired"
    created_date: "When the asset was provisioned"
    last_seen: "Last automated discovery timestamp"

  optional_fields:
    cost_center: "For cost allocation"
    compliance_scope: "SOC2 / HIPAA / PCI / None"
    backup_policy: "Backup schedule reference"
    dr_tier: "Critical / Essential / Standard / Non-essential"
    expiration_date: "For time-limited resources"
    tags: "Key-value pairs from cloud provider"
    dependencies: "Upstream and downstream services"

Scope and evidence limits

Confirm account/subscription/organization and approved regions before queries.

Use read-only discovery identities and restricted output directories: asset data

contains internal addresses, account and owner information. These scripts enumerate

selected services, not every resource. AWS tagging API omits never-tagged resources;

Azure Resource Graph needs pagination beyond its default page; Config sees only

recorded supported resources. Record per-provider permission errors and pagination,

never label a failed query empty or an unseen asset retired. Repeat region-scoped

queries for each authorized region and reconcile a stable account+region+resource

ID before CMDB writes. Config setup below mutates paid account settings and needs

an approved recorder role, delivery channel and change plan first. Tag policies

constrain supported tag values but do not require tags on every resource; provider

default_tags and Config evaluations are not universal preventative enforcement.

The tag-policy YAML is an illustrative wrapper; actual Organizations policy is JSON

with tags at its root, not tag_policy.

AWS Resource Discovery Script

#!/usr/bin/env bash
# aws-asset-discovery.sh - Selected AWS services in the configured region
set -euo pipefail
umask 077
set -o noclobber

OUTPUT_DIR="./asset-inventory/aws/$(date +%Y-%m-%d)"
mkdir -p "$OUTPUT_DIR"
ACCOUNT_ID=$(aws sts get-caller-identity --query Account --output text)

echo "=== AWS Asset Discovery for Account $ACCOUNT_ID ==="

# EC2 Instances
echo "--- EC2 Instances ---"
aws ec2 describe-instances \
  --query 'Reservations[*].Instances[*].{
    InstanceId:InstanceId,
    Type:InstanceType,
    State:State.Name,
    AZ:Placement.AvailabilityZone,
    VpcId:VpcId,
    PrivateIP:PrivateIpAddress,
    PublicIP:PublicIpAddress,
    LaunchTime:LaunchTime,
    Name:Tags[?Key==`Name`].Value|[0],
    Owner:Tags[?Key==`Owner`].Value|[0],
    Environment:Tags[?Key==`Environment`].Value|[0]
  }' --output json | jq 'flatten' > "$OUTPUT_DIR/ec2-instances.json"

# RDS Databases
echo "--- RDS Instances ---"
aws rds describe-db-instances \
  --query 'DBInstances[*].{
    DBInstanceId:DBInstanceIdentifier,
    Engine:Engine,
    EngineVersion:EngineVersion,
    Class:DBInstanceClass,
    Status:DBInstanceStatus,
    MultiAZ:MultiAZ,
    Encrypted:StorageEncrypted,
    Endpoint:Endpoint.Address,
    BackupRetention:BackupRetentionPeriod
  }' --output json > "$OUTPUT_DIR/rds-instances.json"

# S3 bucket names only. Per-bucket location/encryption/versioning need separate
# bounded queries with explicit unknown/access-denied states, not "none" defaults.
aws s3api list-buckets --query 'Buckets[*].{Name:Name,Created:CreationDate}' \
  --output json > "$OUTPUT_DIR/s3-buckets.json"

# Lambda Functions
echo "--- Lambda Functions ---"
aws lambda list-functions \
  --query 'Functions[*].{
    Name:FunctionName,
    Runtime:Runtime,
    MemorySize:MemorySize,
    Timeout:Timeout,
    LastModified:LastModified,
    CodeSize:CodeSize
  }' --output json > "$OUTPUT_DIR/lambda-functions.json"

# VPCs and Security Groups
echo "--- VPCs ---"
aws ec2 describe-vpcs \
  --query 'Vpcs[*].{
    VpcId:VpcId,
    CidrBlock:CidrBlock,
    State:State,
    IsDefault:IsDefault,
    Name:Tags[?Key==`Name`].Value|[0]
  }' --output json > "$OUTPUT_DIR/vpcs.json"

echo "--- Security Groups ---"
aws ec2 describe-security-groups \
  --query 'SecurityGroups[*].{
    GroupId:GroupId,
    GroupName:GroupName,
    VpcId:VpcId,
    Description:Description,
    IngressRuleCount:length(IpPermissions),
    EgressRuleCount:length(IpPermissionsEgress)
  }' --output json > "$OUTPUT_DIR/security-groups.json"

# IAM Users and Roles
echo "--- IAM Users ---"
aws iam list-users \
  --query 'Users[*].{UserName:UserName,Created:CreateDate,PasswordLastUsed:PasswordLastUsed}' \
  --output json > "$OUTPUT_DIR/iam-users.json"

echo "--- IAM Roles ---"
aws iam list-roles \
  --query 'Roles[*].{RoleName:RoleName,Created:CreateDate,LastUsed:RoleLastUsed.LastUsedDate}' \
  --output json > "$OUTPUT_DIR/iam-roles.json"

# EKS Clusters
echo "--- EKS Clusters ---"
aws eks list-clusters --query 'clusters' --output json | jq -r '.[]' | while read -r cluster; do
  aws eks describe-cluster --name "$cluster" \
    --query 'cluster.{Name:name,Version:version,Status:status,Endpoint:endpoint,Created:createdAt}'
done | jq -s '.' > "$OUTPUT_DIR/eks-clusters.json"

# KMS Keys
echo "--- KMS Keys ---"
aws kms list-keys --query 'Keys[*].KeyId' --output text | tr '\t' '\n' | while read -r key_id; do
  aws kms describe-key --key-id "$key_id" \
    --query 'KeyMetadata.{KeyId:KeyId,Description:Description,State:KeyState,Created:CreationDate,Manager:KeyManager}'
done | jq -s '.' > "$OUTPUT_DIR/kms-keys.json"

# Generate summary
echo "=== Inventory Summary ==="
echo "EC2 Instances: $(jq 'length' "$OUTPUT_DIR/ec2-instances.json")"
echo "RDS Instances: $(jq 'length' "$OUTPUT_DIR/rds-instances.json")"
echo "S3 Buckets: $(jq 'length' "$OUTPUT_DIR/s3-buckets.json")"
echo "Lambda Functions: $(jq 'length' "$OUTPUT_DIR/lambda-functions.json")"
echo "VPCs: $(jq 'length' "$OUTPUT_DIR/vpcs.json")"
echo "Security Groups: $(jq 'length' "$OUTPUT_DIR/security-groups.json")"
echo "IAM Users: $(jq 'length' "$OUTPUT_DIR/iam-users.json")"
echo "IAM Roles: $(jq 'length' "$OUTPUT_DIR/iam-roles.json")"

echo "Inventory saved to $OUTPUT_DIR"

AWS Config Rules for Inventory Compliance

# Enable AWS Config recorder
aws configservice put-configuration-recorder \
  --configuration-recorder name=default,roleARN=arn:aws:iam::123456789012:role/aws-config-role \
  --recording-group allSupported=true,includeGlobalResourceTypes=true

# Start recording
aws configservice start-configuration-recorder --configuration-recorder-name default

# Enable required-tags Config rule
aws configservice put-config-rule --config-rule '{
  "ConfigRuleName": "required-tags",
  "Source": {
    "Owner": "AWS",
    "SourceIdentifier": "REQUIRED_TAGS"
  },
  "InputParameters": "{\"tag1Key\":\"Owner\",\"tag2Key\":\"Environment\",\"tag3Key\":\"CostCenter\",\"tag4Key\":\"DataClassification\"}",
  "Scope": {
    "ComplianceResourceTypes": [
      "AWS::EC2::Instance",
      "AWS::RDS::DBInstance",
      "AWS::S3::Bucket",
      "AWS::Lambda::Function"
    ]
  }
}'

# Config rule for encryption compliance
aws configservice put-config-rule --config-rule '{
  "ConfigRuleName": "encrypted-volumes",
  "Source": {
    "Owner": "AWS",
    "SourceIdentifier": "ENCRYPTED_VOLUMES"
  }
}'

aws configservice put-config-ru

… (nội dung đầy đủ có trong gói tải về)