CloudStack relies on two key databases — cloud and cloud_usage.

Overview

The CloudStack Management Server interacts continuously with a database backend. Every API call, VM operation, or event updates these databases.

cloudstack-management
        │
        ▼
   MySQL Server
     ├── cloud
     └── cloud_usage

Database Roles

cloud

The core operational database.
Stores all configuration and runtime state.

Key responsibilities:

  • VM instances, volumes, and templates
  • Networking, IP addresses, and VLANs
  • User accounts, roles, and API keys
  • Async jobs, events, and orchestration logs

cloud_usage

The usage and metering database.
Stores data for reporting, billing, and analytics.

Key responsibilities:

  • Resource usage (CPU, memory, network)
  • Event-driven usage tracking
  • Historical data for metering systems

MySQL Configuration for CloudStack

Before deployment, MySQL must be tuned for high concurrency and transactional safety.

Example configuration (/etc/mysql/mysql.conf.d/mysqld.cnf):

[mysqld]
server_id = 1
sql-mode="STRICT_TRANS_TABLES,NO_ENGINE_SUBSTITUTION,ERROR_FOR_DIVISION_BY_ZERO,NO_ZERO_DATE,NO_ZERO_IN_DATE,NO_ENGINE_SUBSTITUTION"
innodb_rollback_on_timeout=1
innodb_lock_wait_timeout=600
max_connections=1000
log-bin=mysql-bin
binlog-format='ROW'

Why it matters:

  • Prevents silent data corruption
  • Allows long-running transactions
  • Ensures durable writes
  • Supports future replication setups

Initialization

CloudStack includes a setup utility to create and seed both databases:

cloudstack-setup-databases cloud:cloud@localhost --deploy-as=root: -i 192.168.122.10

This command:

  1. Connects to MySQL as root
  2. Creates cloud and cloud_usage
  3. Creates the user cloud@localhost with password cloud
  4. Populates the schema and default data