Distributed Tracing Tracing a user’s network traffic through a monolith is relatively straightforward because a request only has one or two network hops. So, if something fails, then it’s easy to find a stack trace for that user. With a microservice based architecture, however, a single request from a user can spawn many network hops around internal services. Therefore, tracing a user through the system requires navigating loads of different log files and piecing them together. To avoid this headache-inducing exercise, dig into the use of correlation IDs for user requests. Correlation IDs identify a cohesive journey across all internal services and allow distributed tracing tools to monitor these journeys visually.
CONCLUSION The microservices concept continues to gain popularity, showing itself as a clear means to accelerate software evolution. In financial services, the agility itself answers many of the industry’s questions around digital transformation and scalability in the face of constantly changing customer, business, and regulatory requirements. Though firms are likely to experience increased operational overhead during the shift to microservices – particularly in areas such as building, deployment, and monitoring – these challenges can be mitigated over time with the right planning, tools, and patterns. Any architectural change is subject to the needs of a particular organization, but microservices are a sensible choice for financial services firms who are seeking structural reform to solve real business problems. Responsible adoption could mean distinct flexibility for the long-haul.
© 2017 PROVENIR ALL RIGHTS RESERVED
Made with FlippingBook Digital Publishing Software