Understanding the most of the AzureKubernetesServices to design your application

In recent past application has changed his base shell from legacy to containers, containers to micro services. To better utilize your application, one should understand the ecosystem of infrastructure to run the applications.

In this blog we are going to understand the Managed service offering by Microsoft Azure and how better we can utilize the AKS with full potential.

Nature of the application:

This is top priority; one should have the complete understanding to design the infrastructure solution. You application may be write extensive, maybe it requires to run on high compute, maybe you want to run monitoring to get responses from the api server frequently. You might not be sure about the high compute usage timings.

Before designing the infrastructure solution, you must have these answers with you. This will help to build a system which will isolate the potential failure in the user environment.

Designing infrastructure for you production workload with high availability:

Choosing the correct vm size, considering you have already performed the benchmarking of the application this will help you to choose the correct VM size, also make sure you have the correct request and limits set on deployment definition (request is the parameter kube-scheduler will check before electing the correct candidate). You can use kubectl describe node <Node name> to see the allocatable and allocated resources. Also, keep validation the results post deployment of the application. (https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/)

Considering the application is stateless and your queue depth is crossing the mark 10, then you should considering the ephemeral os SSD(consider this as temp disk). https://docs.microsoft.com/en-us/azure/virtual-machines/ephemeral-os-disks some of the key features are :

· Ideal for stateless applications.

· Supported by Marketplace, custom images, and by Azure Compute Gallery (formerly known as Shared Image Gallery).

· Ability to fast reset or reimage VMs and scale set instances to the original boot state.

· Lower latency, similar to a temporary disk.

· Ephemeral OS disks are free, you incur no storage cost for OS disks.

· Available in all Azure regions.

Going with the uptime SLA: This is recommended for the production cluster, Uptime SLA is a tier to enable a financially backed, higher SLA for an AKS cluster. Clusters with Uptime SLA, also regarded as Paid tier in AKS REST APIs, come with greater amount of control plane resources and automatically scale to meet the load of your cluster. (https://docs.microsoft.com/en-us/azure/aks/uptime-sla)