telemetry-go

Telemetry Sample

This sample runs a simple web server that makes calls to other in-cluster services and responds to requests with "Hello World!". The purpose of this sample is to show generating metrics, logs and distributed traces. This sample also shows how to create a dedicated Prometheus instance rather than using the default installation.

Prerequisites

1. A Kubernetes cluster with Knative Serving installed.
2. Check if Knative monitoring components are installed:

kubectl get pods --namespace knative-monitoring

• If pods aren't found, install Knative monitoring component.

1. Install Docker.
2. Check out the code:

go get -d github.com/knative/docs/serving/samples/telemetry-go

Setup

Build the application container and publish it to a container registry:

1. Move into the sample directory:

cd $GOPATH/src/github.com/knative/docs

1. Set your preferred container registry:

export REPO="gcr.io/<YOUR_PROJECT_ID>"

This example shows how to use Google Container Registry (GCR). You will need a Google Cloud Project and to enable the Google Container Registry API.

1. Use Docker to build your application container:

docker build \
  --tag "${REPO}/serving/samples/telemetry-go" \
  --file=serving/samples/telemetry-go/Dockerfile .

1. Push your container to a container registry:

docker push "${REPO}/serving/samples/telemetry-go"

1. Replace the image reference path with our published image path in the configuration file (serving/samples/telemetry-go/sample.yaml):

   • Manually replace:
     image: github.com/knative/docs/serving/samples/telemetry-go with image: <YOUR_CONTAINER_REGISTRY>/serving/samples/telemetry-go

   Or

   • Use run this command:

   perl -pi -e "s@github.com/knative/docs@${REPO}@g" serving/samples/telemetry-go/sample.yaml
   

Deploy the Service

Deploy this application to Knative Serving:

kubectl apply --filename serving/samples/telemetry-go/

Explore the Service

Inspect the created resources with the kubectl commands:

• View the created Route resource:

kubectl get route --output yaml

• View the created Configuration resource:

kubectl get configurations --output yaml

• View the Revision that was created by the Configuration:

kubectl get revisions --output yaml

Access the Service

To access this service via curl, you need to determine its ingress address.

1. To determine if your service is ready:
   Check the status of your Knative gateway:

kubectl get svc knative-ingressgateway --namespace istio-system --watch

When the service is ready, you'll see an IP address in the EXTERNAL-IP field:

NAME                     TYPE           CLUSTER-IP     EXTERNAL-IP      PORT(S)                                      AGE
knative-ingressgateway   LoadBalancer   10.23.247.74   35.203.155.229   80:32380/TCP,443:32390/TCP,32400:32400/TCP   2d

CTRL+C to end watch.

Check the status of your route:

kubectl get route --output yaml

When the route is ready, you'll see the following fields reported as:

status:
  conditions:
    ...
    status: "True"
    type: Ready
    domain: telemetrysample-route.default.example.com

1. Export the ingress hostname and IP as environment variables:

export SERVICE_HOST=`kubectl get route telemetrysample-route --output jsonpath="{.status.domain}"`
export SERVICE_IP=`kubectl get svc knative-ingressgateway --namespace istio-system --output jsonpath="{.status.loadBalancer.ingress[*].ip}"`

1. Make a request to the service to see the Hello World! message:

curl --header "Host:$SERVICE_HOST" http://${SERVICE_IP}

1. Make a request to the /log endpoint to generate logs to the stdout file and generate files under /var/log in both JSON and plain text formats:

curl --header "Host:$SERVICE_HOST" http://${SERVICE_IP}/log

Access Logs

You can access to the logs from Kibana UI - see Logs for more information.

Access per Request Traces

You can access to per request traces from Zipkin UI - see Traces for more information.

Accessing Custom Metrics

You can see published metrics using Prometheus UI. To access to the UI, forward the Prometheus server to your machine:

kubectl port-forward $(kubectl get pods --selector=app=prometheus,prometheus=test --output=jsonpath="{.items[0].metadata.name}") 9090

Then browse to http://localhost:9090.

Clean up

To clean up the sample service:

kubectl delete --filename serving/samples/telemetry-go/