Why ESB Fails

John Spacey, updated on January 14, 2023

Enterprise service bus, or ESB, is an architectural pattern and class of technology that puts an intermediary between data publishers and subscribers to simplify integration. Architecturally this makes sense. Under an ESB model, a publisher publishes its data or services via one interface and subscribers access data and services via one interface. This is significantly less complex than building interfaces between each and every publisher and subscriber that become a complex web of integration. Despite its architectural advantages, ESB can fail. When it does it is typically for one or two reasons:

Single Point of Failure

ESB introduces a single point of failure that may be less resilient than independent custom integrations. Sure, ESB platforms may be deployed in a highly scalable configuration with a large number of instances. However, if you're using a single ESB product that technology is a single point of failure. If a serious bug occurs, all an organization's integration flows may stop. This can become a mess as things queue up, transactions fail and things fall out sync.

Project Dependencies

Projects that can be handled by a single developer or team tend to be fast. With each additional team or set of specialized skills you add to a project, things tend to become slower, more expensive and more likely to fail. ESB is often handled by a special team and may be considered a specialized skill. As such, any project with an integration component requires the ESB team. This can be perceived as an overhead and project bottleneck that may be resisted.

Why ESB Fails

Single Point of Failure

Project Dependencies

Data

Trending