Scaling Solutions with Multiple Dimensions

Sun, Jun 28, 2015

With older monolithic solutions there was a restricted number dimensions that could be used for scaling capacity. Vertical scaling involved allocation of a more and more resources on a single host server. It was not very flexible and there is a finite upper limit. The alternative was to scale horizontally with multiple instances but that too has significant challenges for a monolithic architecture solution.

The following diagram shows how a Microservice Architecture addresses the same challenge.

The extra dimension introduced via the functional decomposition allows much greater flexibility to increase capacity at determined bottlenecks. It is not without some challenges though, as the decentralised data and functional segmentation does mean that updating am existing service with existing service dependencies (e.g. Service A depends on Service B for entity update implies a need to assess impacts of changing the entity structure). Service version control and lifecycle must be considered up front.

Additionally, the distributed service responsibility also means that principles such as “Design for Failure” need to be addressed as part of the Solution Design.

For more discussion see - Introduction to Microservices and The Art of Scale