Top-line service metrics dashboard for linkerd


Dead simple monitoring for linkerd.

linkerd-viz screenshot

linkerd-viz is a monitoring application based on Prometheus and Grafana, autoconfigured to collect metrics from linkerd. linkerd-viz currently supports DC/OS and Kubernetes.

linkerd-viz assumes linkerd has already been deployed onto your cluster, and your applications have been configured to route via linkerd. You'll also need to configure linkerd with the io.l5d.prometheus telemeter, to expose the stats that are displayed by linkerd-viz. You should add this block to your linkerd configuration file:

- kind: io.l5d.prometheus

For more information on getting started with linkerd have a look at our Getting Started guides.

Build Docker image

docker build -t buoyantio/linkerd-viz:latest .

Local boot

docker run -p 3000:3000 -p 9191:9191 buoyantio/linkerd-viz

DC/OS Deploy

Install the official linkerd-viz DC/OS Universe package

dcos package install linkerd-viz

Custom installation

dcos marathon app add dcos/linkerd-viz.json

View dashboard

open $PUBLIC_NODE:3000

Kubernetes Deploy

kubectl apply -f k8s/linkerd-viz.yml

View dashboard

open http://$(kubectl get svc linkerd-viz -o jsonpath="{.status.loadBalancer.ingress[0].*}")

Consul Deploy

First, examine and edit consul/prometheus-consul.yml to fit your ecosystem.

Run the Consul agent locally:

docker run -d --net=host -e 'CONSUL_LOCAL_CONFIG={"leave_on_terminate": true}' consul agent -bind=<external ip> -retry-join=<root agent ip>

for more information see Running Consul Agent in Client Mode.

Boot linkerd-viz locally:

docker run -p 3000:3000 -p 9191:9191 buoyantio/linkerd-viz consul

View dashboard

open localhost:3000