Over the years, we've seen numerous teams making the same mistakes. From accruing several clusters to not providing adequate hands-on time for set up and security. These mistakes can lead to sprawl, costly delays, slow upgrades, and security concerns.
Avoid these mistakes altogether and prioritize velocity and cost savings with Giant Swarm's management cluster.
We deal with the complexity so our customers don't have to. Our goal is to abstract the complexity of running Kubernetes away from our customers.
For every data center or cloud region that we manage, we run a management cluster, which is a Kubernetes cluster. Thus, if the management cluster crashes, it does not become a single point of failure. And your business does not grind to a halt. Within our management cluster, we run operators that automate the lifecycle of our customers’ cloud-native platform.
The common assumption is that there is a slight or perhaps an indistinguishable difference. However, upon closer inspection, this is not the case - at all.
In simple terms, a cluster isn't “operated” by Terraform, Kops, or Kubeadm. These are tools that make it easier to set up a cluster and that’s where it ends. There is no ongoing management. There are no Day 2 operations.
On the other hand, operators are dynamic. They continuously reconcile the actual state of the cluster to its desired state. If something goes wrong, the operator resolves it for you.
With Giant Swarm, you are always within a sliding window of the three latest Kubernetes versions. We also keep you up to date with your cloud-native stack.
Our take on versioning stems from the concept of immutable infrastructure. As such, our releases are immutable. The infrastructure and its components do not change over time for a cluster release. The code that manages the cluster is also immutable meaning it does not change either. If we have to fix the management code, we will also bump a version. This way we make sure we don’t introduce bugs into a working set-up and there are no surprises over time.
As a result, our customers can upgrade on their preferred schedules with confidence. This allows different teams (e.g. development, testing, and production) in your organization to work independently of each other. Thus, removing upgrade paralysis.