Microservices: From System of Record to System of Action

In a previous blog post we took a look at some of the misperceptions that exist with microservices, and also explained the importance of flattening the network, eliminating data silos, and security considerations. In this post, along with accompanying videos, we take a trip down memory lane, and discuss how enterprise systems have evolved from the days of the ERP where data was fairly static, to today’s systems of action.

The three previous eras:

1. System of record - Examples of these systems are the enterprise resource planning application (ERP) which stores an organization’s general ledger, and the customer relationship management application (CRM), which stores sales pipeline and related data. These systems are still critical to enterprise success.
2. System of engagement - These enterprise systems integrate multiple data feeds into a portal to attract and retain prospects, customers, and partners.
3. System of insight - This technology stack support enterprises in their goal of improving their customers’ experience through the consumption, collection, and analysis of data from the combined sources of systems of record and systems of engagement.
    

Watch part 1 of the interview with JJ where he discusses the evolution outlined above.

It should be noted that the systems of the previous eras are not going away. Enterprises always have to balance traditional architectures with new paradigms.

This brings us to today and the system of action, where much of the processing of data can no longer be done by humans looking at different data analysis, but must be done by machines as the data arrives. Enterprises today need to build data pipelines that integrate data from existing systems, along with leveraging new data sources that can be processed in real-time.

Modern Architecture, Powered by Streamlio

Data Layer

The stream storage layer provided by Streamlio, built on Apache BookKeeper, provides the resiliency and scalability needed for a system of action, without incurring much of the cost associated with traditional technologies. Its flexibility means that it can be deployed on premise, in the cloud, or in a hybrid environment. This data infrastructure makes data available wherever and wherever you need it.

Event Layer

Built on Apache Pulsar, Streamlio makes it possible to have a distributed eventing system across data centers and even extend it to hybrid cloud, while minimizing the tradeoffs required of a traditional messaging platform.

Watch part 2 of the interview with JJ where he discusses the requirements needed to build a system of action.

Conclusion

For enterprises looking to adopt a microservice-based approach to application development, it is important to choose the right data and event infrastructure upon which to build real-time applications. Here at Streamlio we are proud to provide the platform for fast data, built on best of breed open-source technologies, Apache BookKeeper, Apache Pulsar, and Heron.

Thanks,
Christian Hasker and JJ Jeyappragash