MySQL for Akash

A production-ready MySQL solution optimized for Akash Network deployments, with robust standalone mode and experimental cluster support.

⚠️ WARNING: Cluster mode is currently a work in progress and should not be used in production environments. Stick to standalone mode for production workloads until cluster mode is fully tested and stable.

Features

• Automatic master/slave failover in cluster mode
• Dynamic configuration optimization based on available resources
• Automated encrypted backups to S3
• GTID-based replication
• Comprehensive health monitoring
• Zero-config standalone mode

Environment Variables

Required

• MYSQL_ROOT_PASSWORD: Root password for MySQL
• MYSQL_REPL_USERNAME: Replication user (required for cluster mode)
• MYSQL_REPL_PASSWORD: Replication password (required for cluster mode)

S3 Backup Configuration (Required for backups)

• S3_ACCESS_KEY: S3 access key
• S3_SECRET_KEY: S3 secret key
• S3_ENDPOINT: S3 endpoint URL
• S3_BUCKET: S3 bucket name
• S3_PATH: Path within bucket for backups
• S3_SSL: Enable SSL for S3 (true/false)

Backup Configuration

• BACKUP_CONFIG_DIR: Backup configuration directory (default: /etc/mysql/backup)
• BACKUP_RETENTION_DAYS: Number of days to retain backups
• BACKUP_FULL_INTERVAL: Interval between full backups in seconds
• BACKUP_INCR_INTERVAL: Interval between incremental backups in seconds

Cluster Configuration

• CLUSTER_MODE: Enable cluster mode (default: false)
• ETCDCTL_ENDPOINTS: etcd endpoints (required for cluster mode)
• ETC_USERNAME: etcd authentication username (optional)
• ETC_PASSWORD: etcd authentication password (optional)
• ETC_PREFIX: etcd key prefix for all MySQL cluster data (default: /mysql)

etcd Schema

The cluster uses etcd for coordination with the following structure:

Base Paths

/mysql/                           # Base path for all MySQL cluster data
    /nodes/                       # Node status and metadata
    /topology/                    # Cluster topology information

Node Status

/mysql/nodes/<node_id>           # Node status and metadata
{
    "status": "online|offline",
    "role": "master|slave",
    "host": "hostname",
    "port": "port",
    "last_seen": "timestamp",
    "gtid_position": "current_gtid",
    "health": {
        "status": "string",
        "connections": number,
        "uptime": number
    }
}

Topology Management

/mysql/topology/master           # Current master reference

Lease Management

• Each node maintains a lease with TTL
• Health updates occur every 5 seconds
• Node status automatically marked offline when lease expires

Backup System

• Encrypted backups using AES256
• Full and incremental backup support
• Automatic backup verification
• Streaming backup support via port 4444
• S3-compatible storage support

Health Monitoring

• Continuous health checks including:
  • Process status
  • MySQL connectivity
  • Read/Write capability
  • Replication status
• Automatic failover on master failure
• Error condition monitoring

Configuration

The system automatically optimizes for:

• Available memory
• CPU cores
• Container limits
• Kubernetes environment detection
• InnoDB settings
• Performance schema (when >8GB RAM available)

License

MIT License - see LICENSE file for details