Localhost Conference 2024_ Building a Flexible and Scalable Database Strategy Across Multiple Clouds
1 like22 views
The document discusses building a flexible and scalable multi-cloud database strategy, emphasizing the characteristics and benefits of deploying databases across multiple cloud environments. It outlines various use cases such as disaster recovery, environment optimization, and control over data, while also addressing the challenges of managing and deploying multi-cloud databases. Practical steps and best practices for deployment, maintenance, and security considerations are also provided.
Localhost Conference 2024_ Building a Flexible and Scalable Database Strategy Across Multiple Clouds
- 1. 1 1 Building a Flexible & Scalable Database Strategy Across Multiple Clouds Divine Odazie, Severalnines Localhost Conference 2024, Virtual <img src="Divine.jpg" alt="Divine" width="300" height="300">
- 2. 2 2 Divine Odazie ● Technology Evangelist at Severalnines ● Electrical/Electronics Engineer ● Certified Kubernetes Application Developer (CKAD) ● Data on Kubernetes Ambassador ● Certified AWS Solutions Architect ● I’m excited to speak at Localhost 2024! Database ops, your way
- 3. 3 1. What is a multi-cloud database? 2. Use cases of multi-cloud database deployment 3. DEMO: Deploying a multi-cloud database 4. Maintaining a multi-cloud database deployment
- 4. 4 Multi-cloud = Multiple different clouds 89% of enterprises operate multi-cloud - Flexera 2023 State of the Cloud Report Database ops, your way
- 5. 5 What is a multi-cloud database? A database deployment that operates across two or more clouds. This could can be on public cloud, private clouds or hybrid (a mixture of both). Any database type is a candidate for multi-cloud deployment - Relational (MySQL, PostgreSQL), - NoSQL (MongoDB, Elasticsearch) - Key-value (Redis), etc Database ops, your way
- 6. 6 Use cases for multi-cloud databases Database ops, your way
- 7. 7 Disaster recovery ● Organizations need to safeguard against major outages. ● The worst that could happen, can happen!! ● Eliminates single points of failure Database ops, your way
- 8. 8 8 Environment optimization ● Improve application performance & latency ● Decrease overhead costs ● Better support mission-critical functionality ● Prices change over time Database ops, your way
- 9. 9 Scaling out Database usage frequency Database uptime duration ● No shortage of specific instance types ● Cheaper option for compute resources ● Cloud bursting: overflow traffic on private cloud to a public cloud Database ops, your way
- 10. 10 10 Control over data Organizations need to: ● Stay ahead of regulatory changes ● Remove vendor and environment lock-in ● Predictable licenses Today, we are seeing the rise of Sovereign DBaaS (Database-as-service). Database ops, your way
- 11. 11 Deploying a multi-cloud database Database ops, your way
- 12. 12 Is deploying and managing a multi-cloud database easy? The short answer is…NO! Database ops, your way
- 13. 13 Using Managed DBaaS vs Compute Instances Cloud specific DBaaS ● Limited interoperability with other cloud providers. ● Replication focus on data import, not scalability. ● There are significant limits in terms of what you can or cannot do. Building with compute instances ● Greater flexibility in inter-cloud connectivity. ● More control over database configuration. ● Option to leverage cloud provider tools or open-source software for a cloud-agnostic setup. vs Database ops, your way
- 14. 14 How do you deploy and manage a database across multiple compute instances on different clouds? Database ops, your way
- 15. 15 Steps to deploying a multi-cloud database Configure cloud credentials, SSH and allow traffic Enable necessary ports: SSH (22), xtrabackup (9999), and MySQL (3306)). Deploy the database on the primary cloud provider Deploy a VPN Choose replication type: Synchronous or asynchronous, or both. Create one or more replica on your secondary cloud provider Launch load balancers (ProxySQL) Deploy monitoring tools (Prometheus, grafana, etc) Database ops, your way
- 16. 16 Manual deployment vs Cloud agnostic tooling Manual deployment ● Set up connectivity between each cloud ● Install the database on each instance ● Configure the replication between the cloud ● Configure Backup/restore across the clouds Agnostic tooling e.g ClusterControl ● Only worry about connectivity between the clouds ● Automate the rest ○ Database installation ○ Replication across each clouds ○ Failover ○ Backups ● Observability from the start ● And so much more!! vs Database ops, your way
- 17. 17 Database ops, your way
- 18. 18 Let’s get started! Database ops, your way
- 19. 19 19 AWS EC2 Instance DigitalOcean Droplet MySQL Replication Primary Replica Database ops, your way
- 20. 20 Maintaining a multi-cloud database Database ops, your way
- 21. 21 21 Key Activities Setup monitoring and alerts Database specific alerts: ● Database health ● Cluster health ● Replication ● Queries etc. Operating system alerts: ● CPU Usage ● RAM or SWAP usage ● Disk usage, etc. Investigating database issues and errors ● Start with the logs ● Check the host system’s health Database ops, your way
- 22. 22 22 Security ● Firewalls and VPN ● Encrypt data at rest and in transit ● RBAC ● Regular audits The Best Practices Checklist Scaling ● Can you database scale? Replication, LB, read/write split. ● Vertical or horizontal scaling ● Automate where possible Performance ● Reduce latency in your cloud connections ● Implement observability ● Automate health checks Governance ● Create a practical governance framework ● Comply to data privacy laws like GDPR, HIPAA Backups and disasters ● Store backups in multiple locations ● Use proper replication ● Automatic backup verification Miscellaneous ● Automated database deployment tools. ● Bring DevOps practice into database deployment Database ops, your way
- 23. 23 Want to try out ClusterControl for free? Other CSPs Contact me [email protected] @Severalnines Severalnines