Skip to content

Latest commit

 

History

History
108 lines (76 loc) · 4.61 KB

File metadata and controls

108 lines (76 loc) · 4.61 KB

A Walkthrough of Monitoring NATS with Prometheus and Grafana

This walkthrough covers the basic setup and configuration for monitoring a NATS server using the NATS Prometheus exporter, Prometheus, and Grafana.

Prerequisites

For this walkthrough, you'll need to install the NATS Prometheus Exporter, NATS Server, Prometheus, and Grafana

NATS Server

go get github.com/nats-io/nats-server

Prometheus

You'll need Prometheus installed; refer to the Prometheus download and installation instructions.

Grafana

For this walkthough you'll also need to download and install Grafana.

Launching the processes

Ensure you are working in the walkthrough directory:

cd walkthrough

It is helpful to start four console windows, one for each component.

1) Start the NATS server and NATS Prometheus Exporter

Be sure to include the a monitoring port:

nats-server -m 8222

Next, launch the NATS prometheus exporter. Here we configure the same monitor port we specified when launching the NATS server. If a NATS server is not running, the exporter will retry to connect to the NATS server indefinitely.

prometheus-nats-exporter -varz -jsz=all http://localhost:8222

You should see something like this:

[21690] 2017/05/02 10:10:37.461404 [INF] Prometheus exporter listening at 0.0.0.0:7777

2) Start Prometheus

prometheus

Prometheus will use the configuration file located in this directory.

You'll see something like this:

INFO[0000] Starting prometheus (version=1.3.1, branch=, revision=)  source=main.go:75
INFO[0000] Build context (go=go1.7.3, user=brew@yosemitevm.local, date=20161105-11:15:15)  source=main.go:76
INFO[0000] Loading configuration file prometheus.yml     source=main.go:247
INFO[0000] Loading series map and head chunks...         source=storage.go:354
INFO[0000] 0 series loaded.                              source=storage.go:359
INFO[0000] Listening on :9090                            source=web.go:240
WARN[0000] No AlertManagers configured, not dispatching any alerts  source=notifier.go:176
INFO[0000] Starting target manager...                    source=targetmanager.go:76

For this walkthrough we use the default URL, http://localhost:9090.

3) Start Grafana

Start your Grafana server. Your settings may differ. Try starting with the server config included in the Grafana tarball.

grafana-server --config=/usr/local/etc/grafana/grafana.ini --homepath /usr/local/share/grafana cfg:default.paths.logs=/usr/local/var/log/grafana cfg:default.paths.data=/usr/local/var/lib/grafana cfg:default.paths.plugins=/usr/local/var/lib/grafana/plugins

Note, your configuration and startup may vary

Look for a log line like this to find the Grafana URL:

INFO[05-02|10:36:04] Server Listening logger=server address=0.0.0.0:3000 protocol=http subUrl=

Connect to Grafana, using the correct host and port. The default Grafana user is admin and the default password is admin.

Add a Prometheus data source with the following parameters:

  • Name: NATS-Prometheus
  • Type: Prometheus
  • Url: http://localhost:9090 (default, unless changed above)

Leave the rest as defaults.

Data Source Image

Next import the NATS dashboard, grafana-nats-dash.json into Grafana, and associate the Prometheus datasource you just created. You should start seeing data graph as follows (note that selecting a view of the last five minutes will be more exciting).

Experiment running benchmarks and applications and you'll see a dashboard like the one below!

Dashboard Image

If you're using JetStream, you can import the JetStream dashboard (grafana-jetstream-dash.json) too:

Dashboard Image


⚠️ Note for Helm Deployments: When deploying the Prometheus NATS exporter via the standard Helm charts, the metrics prefix is overridden from gnatsd & jetstream to nats. You can either simply do a find & replace for these prefixes and replace them in the dashboard, or use the grafana-jetstream-dash-helm.json file (where this has been done already).