At Civo we are always committed to listening to our customers and community. In this post, I would like to highlight some of the great work the team has done to make Civo Kubernetes a go-to choice for enterprises.
We've had various requests from the community and customers who wanted to use NetworkPolicy and a CNI (other than Flannel); we went ahead with Cilium. So now, when you launch a cluster, you get an option to choose between Flannel and Cilium.
You said you wanted to use
serviceType=LoadBalancer for your Kubernetes service, and we heard you. You can now create LoadBalancer service type and get the public IP for your workload (again super fast).
Now, when you launch a default cluster, you have a Traefik NodePort service that is installed and bound to 80/443 hostport so that your cluster DNS works in the same way it used to with the cluster public IP. But as soon as you choose to install any app that has a service of type Loadbalancer, it will get a new IP every time. Let us know in suggestions if you have any additional requests regarding this.
Extended range of node sizes
There are now different node size requirements for different types of workloads. One of the most common customer requests has been to have multiple sizes for the cluster nodes. You can now see different size types for specific scenarios with transparent pricing on the cluster launch page. We have also bumped up the storage per node, another one of our most requested features.
Enterprise users often have a whole team that they need to onboard under a single account. The organization and teams feature was released into General Availability in early 2022. This feature allows different teams to access your account or even group multiple accounts under an organization with different roles. For more info, check this post on using the team functionality.
Firewall Deny can enable you to deny traffic from a particular source when you see a spike in traffic from an unknown source. So you can now allow/deny traffic for both inbound and outbound traffic
Default storage layer
Changing the default storage layer for your clusters has become crucial to further improve reliability. We're happy to announce this change is complete, and for new clusters, the default storage layer is now OpenEBS Mayastor!
We are about to begin migrating all existing clusters over to Mayastor too.
We're really excited to complete this since our testing has proven that it improves stability to the platform and is also superior in performance.
We have upgraded to Kubernetes version 1.22.2 for all new clusters, and for old clusters, you get an upgrade button that will only upgrade the Kubernetes version, not Traefik. However, you cannot use the same ingress objects with apiVersion: extensions/v1beta1 since version 1.22, so you'll need to make those changes before upgrading.
Check out this post for details of all the required changes.
I want to thank our community and customers for supporting and believing in us. We are working full force to deliver an incredible Kubernetes experience for you. We will continue to help the community with initiatives like Civo Academy, the Developer Advocate program, the Civo Startup program.
Stay tuned; you'll see another raft of exciting announcements in time for KubeCon EU 2022 :)