Kubernetes Challenges

As we discussed earlier – K8S is the de-facto orchestration engine for container-based applications. When it was first introduced, the thought was that we will need a few K8S instances (let's say a production, a test and a dev) to house all applications. But that is not what happened. Because of the need of isolation, reduction in interdependency with other teams, application teams started to demand independent instances of K8S Cluster for their application. This created a cluster explosion – as you can see from this research data. The number of K8S cluster per organization are increasing. This trend will continue for the foreseeable future.

Deploying and managing K8S clusters using a DIY approach is not easy. The K8S cluster explosion we discussed earlier, further adds to the complications. For this reason, on-prem and DIY deployments for K8S started to go down and the trend continues this year. Moreover, the choice of where the K8S cluster is deployed is changing. Organizations are adopting multiple public clouds to deploy their clusters in addition to edge locations to cater to the need of cloud native applications to be delivered at edge.

For reasons like OSS, cluster spread, options explosion that we explore later, managing K8S clusters is not trivial. Most organizations site security as their top concern for K8S followed up by in-consistent Deployment and lack of smooth day 2 operations. Organizations are looking for tools and capabilities for these K8S to make their life easier – top examples are data-security, protection, encryption, LCM, monitoring and alerting. These organizations would happily engage with a service provider or partner who would provide these capabilities or tools.

Let's explore the K8S challenges a little further. K8S is a OSS and most of components / functions needed to support a fully functional modern application are OSS. This the global community of developers is your applications team. This has created an explosion of options in the K8S ecosystem. There are multiple ways of implementing a specific need. This forces developers and teams to spend a lot of time on identifying and selecting the options available to the team sometimes creating tool overlap and confusion.

Journey to Cloud-Agnostic Software Development Lifecycle Cafe