At a glance
Leading multinational telecommunications company offers mobile, device sales, fixed-line, and enterprise solutions.
Their legacy PostgreSQL implementation was putting their critical data at risk and limiting their expansion.
They chose YugabyteDB for its scalable, distributed, resilient architecture, and built-in automation.
Telecom Giant Migrates from PostgreSQL to Meet Goal to Connect 1.5M+ IoT Devices in Five Years
This market-leading Japanese telecommunications company offers a wide array of services, including mobile communications, device sales, fixed-line services, and enterprise solutions. They are focused on expanding and modernizing their core telecommunications services to increase resilience and accelerate the delivery of new services.
Their expansion plans include developing a next-generation IoT platform that combines several existing services into a new, cloud-native environment powered by containerized applications. This transformation involves rewriting applications to be Kubernetes-based and deploying a distributed, cloud-native database.
The new tech stack allows them to offer new solutions across a wider range of industries, extending their global IoT business reach. The new IoT platform initially focuses on utility metering devices, with plans to connect approximately 1.5 million devices across hundreds of tenants.
Challenges
Prior to YugabyteDB, the company operated a complex, expensive multi-vendor, multi-platform environment, with the majority of their applications running on PostgreSQL. As their services expanded, they encountered several limitations with Postgres.
- Deployment Limitations: Although the environment will remain on-premises, there was a need to distribute applications and data across multiple regions using Kubernetes.
- Application Downtime: PostgreSQL struggled with high availability, risking user downtime in the event of two simultaneous failures.
- Complex Scalability: The environment could only scale vertically (up), necessitating costly infrastructure upgrades. It also failed to support data distribution across multiple regions.
- Costly Multi-Vendor Management: High operational costs stemmed from the use of multiple vendors and platforms.
Key Database Requirements
The company needed to replace PostgreSQL with a primary database that could manage device information and be accessible from frontend and backend systems. Due to their future expansion plans, they also needed:
- A cloud-native relational database that could easily be deployed on their container-as-a-service platform.
- A highly available and fault-tolerant database that ensured uninterrupted service during simultaneous Kubernetes (K8) pod failures.
- A horizontally scalable database that could accommodate the connection of additional IoT devices.
- A geographically distributed database to support their future plans for an advanced distributed computing infrastructure and disaster recovery (DR) configurations across two regional data centers in Japan.
- A 100% open source database to avoid vendor lock-in.
- A cost-effective solution that could fulfill the above criteria.
Business Results
10k
operations per second
1.5M+
planned IoT devices
100%
Apache open source
Yugabyte Solution
After a thorough evaluation, the telecom leader selected YugabyteDB as their distributed database of choice based on several key factors:
- YugabyteDB best met their critical database requirements compared to other distributed databases.
- YugabyteDB Anywhere (our self-managed DBaaS solution) provided built-in automation, simplifying Day2 operations and ensuring required levels of observability.
- YugabyteDB has a proven track record of powering large-scale environments at other industry-leading companies.
- Strong implementation support by Yugabyte’s professional services team, including technical guidance on architecture design, deployment, Day 2 operations, and enablement.
Data is distributed across nodes to support ongoing operations during multiple component failures.
Data and applications are seamlessly distributed across on-premises data centers in two Japanese regions.
Simplified and automated Day 2 operations and observability thanks to YugabyteDB Anywhere.