In the last months, I was looking for a solution to manage Kubernetes at scale in multi-cloud.
You have many products in the market like:
1- Gardener is open source from SAP
The gardener has three tiers of clusters:
Garden Cluster top-level management cluster, responsible for:
- Managing the seed clusters and through them managing the shoot clusters.
- Interacting with end-users, allowing them to deploy seeds and shoots.
Seed Cluster: is responsible for managing one or more shoot clusters.
Generally, you have one seed cluster per cloud provider per region because of the latency issues.
Shoot Cluster: deploy your normal workloads: web servers, databases, machine learning. You will have many shoot clusters deployed all over the world across many providers and locations.
- Open source
- Support many cloud providers (AWS, Alibaba Cloud, Azure GCP, Metal-Stack, OpenStack, VMware vSphere)
- Saving cost: in the large-scale clusters, the concept of seeds will save cost. For example, if you have more than 1000 Kubernetes clusters that mean you need to have 3000 Master nodes (control plane) but if you use seed can be reduced to 1000 nodes as seeds can manage multiple shoots.
- Nightmare for setup, especially for upgrades.
- Lack of documentation.
- No enterprise support.
2- Anthos from Google Cloud Platform (GCP):
- Enterprise support.
- One dashboard for all Kubernetes clusters.
- Not open source.
- Not cloud agnostic.
- High cost as you pay per CPU for all nodes in the cluster.
3- Rancher is open source from SUSE
- Open source.
- Easy to setup.
- Good documentation.
- Community and enterprise support.
- Continuous Delivery.
- Some of the API documentation is missing, which I figured out by seeing the request from the web portal.
Top comments (0)