Deploying on Kubernetes

Using the API Deployer, you can deploy your API services to a Kubernetes cluster.

Each API Service Deployment (see Concepts) is setup on Kubernetes as:

• A Kubernetes deployment made of several replicas of a single pod

• A Kubernetes service to expose a publicly available URL which applications can use to query your API

• Setting up
  • Prerequisites
  • Limitations
  • Build the base image
  • Create the Kubernetes infrastructure in DSS
• Deployment on Google Kubernetes Engine
  • Setup
    • Create your GKE cluster
    • Prepare your local docker and kubectl commands
    • Setup the infrastructure
    • Deploy
  • Using GPUs
    • Building an image with CUDA support
    • Create a cluster with GPUs
    • Add a custom reservation
      • At the infrastructure level
      • At the deployment level
    • Deploy
• Deployment on Azure AKS
  • Setup
    • Create your ACR registry
    • Create your AKS cluster
    • Prepare your local az, docker and kubectl commands
    • Setup the infrastructure
    • Deploy
  • Using GPUs
    • Building an image with CUDA support
    • Create a cluster with GPUs
    • Add a custom reservation
      • At the infrastructure level
      • At the deployment level
    • Deploy
• Deployment on AWS EKS
  • Setup
    • Create your EKS cluster
    • Prepare your local aws, docker, and kubectl commands
    • Setup the infrastructure
    • Deploy
  • Using GPUs
    • Building an image with CUDA support
    • Create a cluster with GPUs
    • Add a custom reservation
      • At the infrastructure level
      • At the deployment level
    • Deploy
• Deployment on Minikube
  • Setup
    • Create the base image
    • Start DSS with proper env
  • Configure infrastructure
• Managing SQL connections
  • Configuring the connection used for storage of bundled data
  • Configuring the “referenced” connections