2.2 KiB
Neolink in Kubernetes
Introduction
This README provides guidance on deploying the NeoLink application on Kubernetes using the provided manifest files. The deployment is designed to set up a NeoLink application within its own namespace, with a specific configuration for node affinity, resource limits, and service exposure.
Prerequisites
- Kubernetes Cluster
- Command-line tool (kubectl)
- Basic understanding of Kubernetes concepts (Pods, Services, Deployments, ConfigMaps)
Deployment Overview
The deployment consists of four main parts:
-
Namespace Creation: A new namespace neolink is created to isolate the resources.
-
Deployment Configuration: The neolink-app-deployment is set up with:
- A single replica.
- Affinity to arm64 nodes (Though this is optional).
- Resource limits and requests for CPU and memory.
- A custom command and args to use a specific configuration file.
-
Service Exposure: A LoadBalancer type service named neolink exposes the application on port 8554.
-
ConfigMap: Contains the application configuration neolink.toml.
Steps to Deploy
kubectl apply -f manifest.yaml
Configuration Details
- The neolink.toml file in the ConfigMap provides the application configuration. Customize it as per your requirements.
- The deployment is set to favor arm64 architecture nodes via a
nodeAffinity. This is a personal preference as in my cluster all my raspberrpis (except one) run 32bit raspian OS. And to try and get more bang for my buck I was to run neolink on my arm64 8Gb node. Again, this is totaly optional for you. - Resource limits are set to ensure efficient use of cluster resources.
- The LoadBalancer service will make the application accessible externally.
Troubleshooting
- If the pod fails to start, check the pod logs and describe the pod for more details.
- Ensure that the arm64 node is available and schedulable in your cluster (if you're using nodeAffinity).
- Verify that the ConfigMap is correctly applied and mounted.
Conclusion
This guide provides a step-by-step approach to deploying the NeoLink application on a Kubernetes cluster. Modify the manifests as necessary to fit your specific requirements and environment.