philjhale / AzureObservabilitySampleApp

The purpose of this repository is to evaluation Azure Monitor's ability to monitor Function Apps and APIs deployed in Azure Kubernetes Service.

Azure observability sample app

The purpose of this repository is to evaluation Azure Monitor's ability to monitor Function Apps and APIs deployed in Azure Kubernetes Service.

Prerequisites

All

• .NET Core SDK 2.1
• An Azure account

Mac

• Download and install Docker for Mac
• Preferences -> Kubernetes -> Enable Kubernetes

Windows

• You're on your own. Probably Docker and Minikube

Setup steps

Create a resource group

Note, you may want to alter the names of all the Azure resources created.

az login
az group create --name=observability-sample --location=ukwest

Create and publish a Function App

A Function App requires a storage account. The storage account name must be less than 24 characters and only contains numbers and letters.

az storage account create --name observabilitystorage --resource-group observability-sample --sku Standard_LRS
az functionapp create --resource-group observability-sample --storage-account observabilitystorage --name ObservabilitySamplefunctions --runtime dotnet --consumption-plan-location ukwest
func azure functionapp publish GetDogImage

Create an Application Insights resource

At the time of writing this can only be done via the Azure Portal. In the Azure Portal go to "Create a resource" then Application Insights, and hit Create.

• Application Type - General
• Use existing resource group

Once deployed go to the Overview tab and copy the Instrumentation Key. Switch to the Functional App and add a new app setting with the name APPINSIGHTS_INSTRUMENTATIONKEY. Next, because we're using Application Insights we no longer need the built in logging. To disable this, delete the AzureWebJobsDashboard app setting. Hit the Save button at the top of the page. Copy the Instrumentation Key into the Kubernetes deployment file here.

Create an Azure Kubernetes Cluster and deploy a web API

Create the cluster.

az aks create --resource-group observability-sample --name observability-aks-cluster --node-count 1 --enable-addons monitoring --generate-ssh-keys

Install the Azure Kubeternetes Service Command Line Interface if you have not already.

az aks install-cli

Configure kubectl to connect to your cluster.

az aks get-credentials --resource-group observability-sample --name observability-aks-cluster

Deploy the API.

kubectl apply -f deployment.yaml
# or in the WebApi folder
./kubectl-apply.sh

Wait for public IP

kubectl get service observability-sample-service --watch

Once the public IP appears you can access the API in your browser. https://[[PublicIP]]:4000/api/dog