High Availability

Replica distribution

Knovvu applications utilize Kubernetes features to remain operational and accessible even in the event of failures or disruptions. Applications are distributed across different virtual machines and are configured to be replicated in case of application or virtual machine failures. This brings self-healing capabilities, ensuring failures are handled gracefully.

Note that the effectiveness of high availability relies on the resilience of the underlying data center infrastructure supporting Kubernetes. For example, if the virtual machines on which applications are configured to be replicated on share similar infrastructure components (e.g., running on the same hardware), it can negatively impact the availability of Knovvu applications on Kubernetes.

Multi control plane setup

Kubernetes control plane manages the worker nodes of the cluster, where the applications run. As the control plane components are highly critical for the operation of the cluster, when Sestek deploys Kubernetes in production environments (for details refer to Deployment Approaches), control plane has a minimum of three virtual machines with replicated components.

Load balancing

Network traffic is efficiently distributed across multiple instances of Knovvu applications through various load balancing mechanisms, encompassing both external load balancing outside the Kubernetes cluster and internal load balancing within the Kubernetes cluster. This prevents bottlenecks and enhances the overall performance and resilience of Knovvu applications.

Next topic: FAQ