Building and deploying microservice applications

How, when, and why to migrate from a monolithic to a microservice architecture

Download white paper

Case study: Improve operations and lower costs

Company overview

A global asset management firm identified a collection of applications, each with a long history of serving many types of users at the company. Multiple distributed data sources proved to be inconsistent, inaccessible, and a significant hindrance to gaining insights and making informed decisions.

Opportunity

The firm’s data governance team soon recognized the need to craft a definitive data strategy and develop a new data management architecture to prevent teams from creating isolated pockets of data. They turned to the Devbridge team to build a data strategy, develop design patterns for a microservice architecture, and create a plan to centralize access to business data that would start with a pilot rollout of specific services.

Approach

Employing product development best practices, Devbridge designed and implemented a pilot initiative to validate the chosen data management strategy. The plan covered the effort necessary to centralize a subset of the data sources into the master services environment and perform the complex task of reintegrating all dependent applications.

Solution

The team set specific, measurable, and quantifiable goals to validate the investment. A precise set of requirements was developed for speed of data access, resiliency, and scalability for future needs. Kubernetes was chosen to manage containerized deployment, automate operations, and scale microservice performance as necessary.

Results

Post-implementation access speeds exceeded expectations: in complex data retrieval operations, more than 250,000 fields were returning at a rate of 0.0016 milliseconds per data field. By implementing a microservice architecture, the asset management firm reduced ongoing operating costs and decreased the complexity of the data management application. Internal teams were able to access source data reliably and with highly consistent results. The same APIs were exposed to external clients to provide an alternative to their legacy solutions. The initial release shipped to production in under three months, which delighted the client.

Continue to:Know how, when, and why to use microservices