feat: In Kubernetes We Trust

[ibuziuk]

Preview link

https://pr-check-70-che-blog.surge.sh/2024/11/26/@ilya.buziuk-in-kubernetes-we-trust.html

[dkwon17]

Suggested change

Do NOT manage the Kubernetes clusters manually otherwise you would end up with a snowflake environment. Application definitions, configurations, and environments should be declarative and version controlled. Application deployment and lifecycle management should be automated, auditable, and easy to understand. Using a GitOps CD solution for Kubernetes such as link:https://argo-cd.readthedocs.io/[Argo CD] is a must-have when managing a complex application platform for developers.

Do NOT manage Kubernetes clusters manually otherwise you would end up with a snowflake environment. Application definitions, configurations, and environments should be declarative and version controlled. Application deployment and lifecycle management should be automated, auditable, and easy to understand. Using a GitOps CD solution for Kubernetes such as link:https://argo-cd.readthedocs.io/[Argo CD] is a must-have when managing a complex application platform for developers.

[dkwon17]

Suggested change

Although autoscaling is a powerful Kubernetes feature, you cannot always fall back on it, and should always consider predictive scaling by analyzing the load data on your environment to detect daily or weekly usage patterns. If your workloads follow some pattern, e.g. there are huge spikes based on the time of the day, you should consider provisioning worker nodes in advance (e.g. a lot of users turn on their smart speakers in the morning between 7 - 9 am, and there is a huge spike in the requests that on infrastructure level is predicted and handled in advance).

Although autoscaling is a powerful Kubernetes feature, you cannot always fall back on it, and should always consider predictive scaling by analyzing the load data on your environment to detect daily or weekly usage patterns. If your workloads follow some pattern, e.g. there are huge spikes based on the time of the day, you should consider provisioning worker nodes in advance (e.g. a lot of users turn on their smart speakers in the morning between 7 - 9 am, causing a huge spike in requests which can be predicted and handled in advance on the infrastructure level).