Node Pool Resource Allocation and Cluster Creation Optimization in Kubernetes Clusters with CloudStack

[RepoBreather]

I’ve been working with Kubernetes clusters in a CloudStack environment, and I had a couple of questions regarding node pools and cluster creation times. I’m hoping to get insights from the community or contributors who have experience with these topics.

1. Node Pool Resource Allocation

Is it possible to create a node pool in a Kubernetes cluster managed by CloudStack, where specific nodes within the cluster can be assigned certain resources (e.g., compute, memory, storage) from cloudstack console or through CloudStack APIs?
I understand that Kubernetes allows taints, tolerations, and labels to control workload placement, but I’m specifically looking for ways to predefine resource allocation at the node pool level when provisioning nodes. Ideally, I’d like to:

• Define resource specifications for each node pool.
• Use CloudStack APIs to automate the creation of these resource-specific node pools.

If this is possible, could someone guide me on how to achieve this using CloudStack APIs? Any example API calls or configuration snippets would be highly appreciated.

2. Cluster Creation Time Optimization

When I create a Kubernetes cluster with 20 or more nodes, I’ve noticed that while the instances for all nodes are up and running within 5 minutes, the cluster itself takes an additional 10 to 15 minutes before it becomes fully operational.
I suspect that the additional time might be due to:

• CloudStack setting up internal configurations for Kubernetes.
• Delays in networking, provisioning, or control-plane readiness.

Is there any way to reduce the overall cluster creation time in a CloudStack-managed Kubernetes environment? I would love to hear if anyone has optimized their cluster deployment workflows or has specific suggestions to speed up the process.

[RepoBreather]

Hi, can anyone please help with the above issue, is this feature of node pool supported currently or will be in any next release.

[weizhouapache]

@RepoBreather
does #9102 solve your issue ?