In the context of Kubernetes cluster orchestration, managing different types of workloads, or “node groups”, on the same cluster, would require managing several logical entities. On the AWS cloud, for example, that means configuring and managing several ASGs. The same would also be true when it comes to running your worker nodes on spot instances, as spot instance size diversification would again require multiple ASGs as per AWS best practices.
This issue is addressed by one of the key features of Ocean, custom Launch Specifications. These allow you to configure multiple workload types on the same Ocean Cluster. As part of those launch specs, you can configure different sets of Labels and Taints to go along with a custom AMI, User Data script, Instance Profile, Security Group and Root Volume size, which will all be used for the nodes that serve your matching pods. This ensures the ability to run any type of workload on the same Ocean Cluster.
Today, we are happy to announce the addition of instance tags to the list of parameters configurable via Launch Specifications.
This addition will bring advantages such as:
- Allowing customers to enjoy more granular control over their instances.
- Easier debugging, identification and separation of instances in the cloud account to their respective launch specifications.
- The ability to migrate over equivalently tagged workloads such as KOPS instance-groups, with their tags intact.
Custom tags per Launch Specification are available now via API, with UI support arriving soon.
Read more here.