Skip to content

Latest commit

 

History

History
65 lines (48 loc) · 2.37 KB

deployment.md

File metadata and controls

65 lines (48 loc) · 2.37 KB
Raw

Deployment of the Evaluation Cluster

Any cluster can be used for the evaluation. However, the benchmark was developed primarily using kind. Thus, kind is recommended for local deployments.

The resources can be deployed and removed manually using kubectl. However, for more comfort Skaffold is supported to deploy and delete all the benchmark resources.

🐣 Setup

  1. Make sure that all manifests have been generated and placed in the manifests folder

  2. Create a cluster

    • for kind the following command can be used: 
      kind create cluster --config=cluster.yaml

  3. Ensure that kubectl context is set to the target cluster

    • check the current context with: 
      kubectl config current-context
    • the context can be switched with: 
      kubectl config use-context <context-name>

  4. Deploy the benchmark resources by running

    skaffold run

    To avoid waiting for the resources to stabilize add the --status-check=false flag.

    💡 Note: the resources don't have to be in running state in order for the tools to analyzes them.

💀 Teardown

  1. Delete all deployed benchmark resources by running

    skaffold delete

  2. Verify that all the generated resources have been removed by viewing all resources with the kalm-benchmark label

    kubectl get all -A -l app.kubernetes.io/part-of=kalm-benchmark

    💡 Note: It may take a while until all resources have been deleted - please by patient.

  3. Delete the cluster

    • for kind is following command can be used: 
      kind delete cluster --name kalm-benchmark

🆘 Help: Resource is stuck in terminating state

Occasionally, it happens that a resource gets stuck in terminating state. If this happens either forcuflly kill the pod/container/resource or delete and re-create the cluster. See the following links for more details: