Kubernetes consulting and operations: clusters your team understands
Kubernetes plays to its strengths when many services, several teams or strongly fluctuating load come together. Sortia Labs works out with you whether a cluster pays off for your applications, sets it up as traceable code and, if you wish, runs it with updates, monitoring and backups. If a simpler setup is enough, we will tell you.
- Kubernetes consulting
- Kubernetes operations
- Managed Kubernetes
- OpenShift
- Helm charts
- Container platform
- GitManifests and Helm charts
- PipelineBuild and test the image
- ClusterDeployments, ingress, secrets
- MonitoringMetrics, logs, alerts
What we take care of for your Kubernetes setup
Fit check before the first cluster
We look at your services, your team and your load and compare Kubernetes with simpler routes such as Docker Compose. You get a reasoned recommendation, including the effort for setup and ongoing operations.
Cluster setup as code
Managed Kubernetes on AWS (EKS) or Azure (AKS), OpenShift or a cluster on your own servers. We describe network, storage, permissions and ingress with Terraform and Helm, so that every state can be traced and rebuilt.
Getting applications ready for containers
We write Dockerfiles and Helm charts, keep configuration and secrets out of the image and define resources, health checks and restart rules. Testing then runs the same image that later goes to production.
Rollouts with GitOps
With ArgoCD every change reaches the cluster from git. New versions are rolled out step by step, a faulty one can be reset to the previous state, and the history shows who changed what and when.
Databases in or next to the cluster
Operators for PostgreSQL and MySQL, persistent volumes and backups stored outside the cluster. Whether a database belongs in the cluster or runs better beside it, we decide with you based on load and operating effort.
Updates, permissions and capacity
We keep Kubernetes versions, nodes and add-ons up to date and grant permissions on a least-privilege basis. Metrics and alerts show how busy the cluster is and where bottlenecks are building up.
Four steps to a cluster that carries the load
- 1
Assessment
We record services, dependencies, load peaks and your requirements for privacy and availability. From that it becomes clear whether Kubernetes, managed Kubernetes or a leaner setup fits.
- 2
Pilot with one service
One selected service moves into the cluster first, with monitoring, backups and a deployment pipeline. Your team sees on a real example what daily work with Kubernetes looks like.
- 3
Step-by-step adoption
Further services follow one at a time. The current setup keeps running until the cluster has proven itself under real load, and every step has a prepared way back.
- 4
Operations or handover
We keep looking after the cluster or train your team until it handles updates, troubleshooting and deployments itself. Runbooks and documentation stay with you.
Questions about Kubernetes
When does Kubernetes pay off?
Typical signs are many services that are deployed independently, several development teams and load that changes a lot over the day. If a handful of services runs with steady load, a leaner setup is usually cheaper to operate. We compare both routes using your real numbers.
Managed Kubernetes or our own cluster?
EKS on AWS or AKS on Azure take the control plane off your hands, but tie you more closely to the provider and its pricing. Your own cluster on servers in Germany gives you more control but needs more operating work. What usually decides it is who looks after the cluster day to day.
Can you take over an existing cluster?
Yes. We first review versions, permissions, resources, backups and deployments and record what is documented and what is not. You then get a list of risks and a plan for the order in which we address them.
OpenShift or Kubernetes: what is the difference?
OpenShift is the Kubernetes platform from Red Hat, with extra tools for builds, permissions and a web console, and with licence costs. If your company already uses OpenShift, we work on it. For a new cluster we compare both options on cost and operating effort.
Read on.
Is Kubernetes right for your applications?
Tell us briefly which services you run and who looks after them today. You get an honest assessment of whether a cluster makes sense and what the path there looks like.