Integration plays an increasingly important role in IT, connecting applications and APIs, spanning cloud services and internal networks, automating workflows and delivering that most sought-after quality for daily operations: agility.
As integration workloads increase, performance and scalability become critical issues. Organizations can’t afford to have integration processes hit bottlenecks because of limitations in CPU power, memory, or network bandwidth.
In integration, as in other areas of IT, performance matters. IT organizations should consider how they’re going scale integration processes to meet the needs of business users and the workflows they depend on.
Scalability in the Boomi Platform: Atoms, Molecules, and More
Fortunately, we designed Boomi’s integration platform with scalability in mind. To ensure that Boomi software always scales to meet performance requirements, we offer a variety of Boomi configurations for IT architects and integration managers to choose from.
The simplest configuration is Boomi’s cloud service. If all integration endpoints are in the cloud, you can run on a Boomi Atom Cloud.
If one or more endpoints is running on an internal network, Boomi offers a choice of configurations to address the requirements of any integration, whether it’s a straightforward point-to-point integration or a complex workflow involving multiple endpoints and APIs.
Let’s start with the basic building block of the Boomi platform: The Atom.
Boomi Atoms are lightweight Java engines that you can deploy in the cloud or on-premise. Each Atom runs in a single Java virtual machine (JVM). The Atom’s performance is bound by the capacity of its host computer. Most Atoms can handle general integration workloads and even real-time integration workloads easily.
For heavy workloads, especially those involving extract, transform and load (ETL) processes, an IT architect or integration designer might want to choose a more distributed configuration. Boomi Molecules expand the single-tenant Boomi environment to run on one or more machines. Like Atoms, Molecules run on JVMs, but they can span machines, as long as those machines can share a file system using the NFS protocol or Windows File Share.
For heavy workloads, Molecules offer several advantages over Atoms running on their own, including better scalability and fault tolerance.
Want even more scalability? Consider a Boomi Atom Cloud. Running on multiple machines in a multi-tenant environment on Windows or Linux/Unix, an Atom Cloud comprises one or more primary Java programs with many secondary Java runtimes. The Atom Cloud scales across machines and JVMs, creating a highly scalable architecture for integrations and workflows.