Cosmiqube

Service

Kubernetes setup, deployment and ongoing cluster management

Cluster design, GitOps delivery and day-2 operations for EKS, AKS, GKE and self-managed Kubernetes.

Kubernetes is straightforward to install and genuinely hard to operate. The gap shows up months later, in upgrade paralysis, mysterious evictions and a cluster nobody wants to touch. We set clusters up so that the second year is as uneventful as the first week.

What the engagement covers

Cluster architecture

Node pool topology, networking and CNI choice, storage classes, and a sizing model that matches your actual workload profile rather than a template.

GitOps delivery

Argo CD or Flux, so the cluster state is whatever the repository says it is. Rollbacks become a revert, and drift becomes visible instead of invisible.

Autoscaling

Horizontal pod autoscaling with sane metrics, cluster autoscaler or Karpenter for nodes, and requests and limits set from measured usage.

Cluster security

RBAC that reflects real roles, network policies, admission control, image scanning in the pipeline, and secrets sourced from a real secret store.

Upgrades

A tested, repeatable upgrade path for the control plane, nodes and add-ons — the single most common reason teams end up stranded on an unsupported version.

Day-2 operations

Monitoring, alerting, incident response and on-call support, with runbooks written for whoever is holding the pager.

Tools we work with

Not an exhaustive list, and not a claim of certification — just what we reach for most often on this kind of work.

  • Kubernetes
  • EKS
  • AKS
  • GKE
  • Helm
  • Argo CD
  • Flux
  • Karpenter
  • Istio
  • Cilium
  • Prometheus

Common questions

Do we actually need Kubernetes?
Often not. If you run a handful of services with steady traffic, a managed container runtime is cheaper and easier to operate. We will say so — recommending Kubernetes you do not need is a good way to lose a client in year two.
Can you take over a cluster someone else built?
Yes. We begin with an audit covering version support, security posture, resource configuration and upgrade risk, and give you a prioritised remediation plan before we change anything.
Do you offer on-call cover?
Yes, on a retainer, with response targets agreed in writing. We keep the number of clusters we cover deliberately small so the commitment stays real.

Talk to us about kubernetes

Send over what you are dealing with. We will reply with a straight assessment of whether we are the right team for it.

Get in touch