Over the past few years, containerization has evolved at an exponential rate. They can be spinned up or down within a matter of seconds. Both resource-wise and performance-wise, they have been way ahead of the legacy virtual machines. We understood the power of containers and designed our Integration Platform to run on top of containers. Behind the scenes, Kubernetes framework, an open-source container orchestration project introduced by Google, does the wiring of these containers. We also support Red Hat’s OpenShift Container Platform.
With the speed of containers, you can easily start an ESB cluster with any number of instances within a few seconds. All you have to do is to define the cluster, add one or more projects to it, and press the Deploy button. IPS does rest for you.
Clusters created by IPS are auto-healing, meaning that the number of instances is guaranteed to remain the same. In case some instances fail, the platform automatically spawns new instances to compensate for them.
When a new configuration needs to be installed into a working ESB cluster, IPS takes the responsibility of distributing new configurations to target ESBs while maintaining consistency across the whole cluster. All it takes is just a few mouse clicks in the IPS management dashboard.
You can monitor the health and other functional statistics of each ESB cluster from the IPS management dashboard. If you want to go deeper to analyze the performance of each ESB in the cluster, you can easily do that as well.
Sometimes, the number of instances running in a particular cluster may not be the optimum value to serve its purpose. In such cases, the cluster can be easily scaled horizontally to get just the right number of instances up and running. All you have to do is to change the cluster size and press Refresh Cluster, and IPS will take care of the rest.
IPS runs as a distributed system on top of the Kuberenetes framework. If the platform need more computing power, new machines can be added to it, with zero downtime on existing components. Similarly, it is also possible to remove existing machines, and the platform will balance the load across other machines.
Although IPS performs complex stuff to serve users’ need and provide a better quality of service, users do not directly see them. All the complexities are transparent to the user and all he can see is a scaled ESB cluster which integrates his entire system.
If you have your own cluster of machines or an on-premise data center, IPS can utilize those resources to serve as your local integration platform. If you want to migrate your integration services to a public cloud such as AWS, IPS can support that as well. In the future, we also expect to provide support for private cloud hosting services such as OpenStack.