Maria DB cluster:
High availability DB Cluster
The company
The customer is an international company and the largest Ukrainian telecommunications operator, serving millions of mobile subscribers and home Internet subscribers. It provides services using a wide range of mobile and fixed technologies, including 4G, mobile financial services, Big Data, Cloud solutions, cyber protection services, digital TV, and industrial IoT.
The company’s mission is to help people stay connected in the digital world, providing them with reliable and secure technology. The company is committed to providing a safe and secure online experience for its customers, which means taking steps to protect their personal data. The company’s policy emphasizes the importance of customer privacy and security of stored information. It employs a variety of security measures – both physical and digital – that are designed to ensure its customers’ data is kept safe from unauthorized access or misuse.
The client chose our team as it is highly trained and knowledgeable, so we can provide the necessary assistance to ensure a satisfactory customer experience.
Initial situation
The customer has a Galera cluster with 3 nodes, which was built on the Community version of MariaDB.
With the increase in DB size and the lack of balancing between nodes, the speed of the system began decreasing and its’ functionality declined.
Business needs:
- Improve fault tolerance and add geo-redundancy without changing the DB
- Improve productivity
- Balancing requests to the DB
- Create a single entry point for applications and users
The solution
To address the customer's requirements, it was decided to use the licensed version of MariaDB Enterprise Edition and Maxscale. The following steps were taken:
Three data centers were created, with three servers in each, based on geographic locations.
The MariaDB Enterprise Edition was deployed, providing advanced features and performance enhancements.
Two Maxscale nodes (master|slave) were added and managed by Keepalived for load balancing and fault tolerance.
Outcomes
- The customer received a load balancer and read replicas through Maxscale, resulting in increased performance and responsiveness of the system.
- By unifying data centers based on geographical locations, the customer achieved additional fault tolerance and geo-redundancy.
- The Triangu solution successfully addressed the initial productivity challenges faced by the telecom company, improving system speed and functionality.
