Gnatsd

Agent Check

Supported OS:

Overview

Get metrics from gnatsd service in real time to:

Visualize and monitor gnatsd states
Be notified about gnatsd failovers and events.

Setup

Installation

To install the Gnatsd check on your host:

Install the developer toolkit on any machine.
Run ddev release build gnatsd to build the package.
Download the Datadog Agent.
Upload the build artifact to any host with an Agent and run datadog-agent integration install -w path/to/gnatsd/dist/<ARTIFACT_NAME>.whl.

Configuration

Edit the gnatsd.yaml file to point to your server and port, set the masters to monitor

host: set to the gnatsd host to monitor
port: set to the monitoring port used by gnatsd
tags: add these tags to recorded metrics
server_name: set to what should be displayed in DataDog

Validation

When you run datadog-agent info you should see something like the following:

Checks
======

    gnatsd
    -----------
      - instance #0 [OK]
      - Collected 23 metrics, 0 events & 1 service checks

Compatibility

The gnatsd check is compatible with all major platforms

Data Collected

Metrics

gnatsd.connz.connections.in_bytes (count)	The number of bytes incoming shown as byte
gnatsd.connz.connections.in_msgs (count)	The number of messages received shown as unit
gnatsd.connz.connections.out_bytes (count)	The number of bytes outgoing shown as byte
gnatsd.connz.connections.out_msgs (count)	The number of messages sent shown as unit
gnatsd.connz.connections.pending_bytes (gauge)	The number of bytes pending ack on a subscription shown as unit
gnatsd.connz.connections.subscriptions (gauge)	The number of subscriptions on a connection shown as unit
gnatsd.connz.num_connections (gauge)	The number of current connections to the NATS broker shown as unit
gnatsd.connz.total (count)	The number of connections ever to the NATS broker shown as unit
gnatsd.routez.num_routes (gauge)	The number of routes in the cluster shown as unit
gnatsd.routez.routes.in_bytes (count)	The number of bytes incoming shown as byte
gnatsd.routez.routes.in_msgs (count)	The number of messages received shown as unit
gnatsd.routez.routes.out_bytes (count)	The number of bytes outgoing shown as byte
gnatsd.routez.routes.out_msgs (count)	The number of messages sent shown as unit
gnatsd.routez.routes.pending_bytes (gauge)	The number of bytes pending shown as unit
gnatsd.routez.routes.subscriptions (gauge)	The number of subscriptions on a connection shown as unit
gnatsd.varz.connections (gauge)	The number of connections to the NATS broker shown as unit
gnatsd.varz.in_bytes (count)	Amount of traffic sent into the cluster shown as byte
gnatsd.varz.in_msgs (count)	The number of messages passed into the cluster shown as unit
gnatsd.varz.mem (gauge)	Amount of memory currently used by the process shown as byte
gnatsd.varz.out_bytes (count)	Amount of traffic sent from the cluster shown as byte
gnatsd.varz.out_msg (count)	The number of messages sent out of the cluster shown as unit
gnatsd.varz.remotes (gauge)	The number of currently connected remotes shown as unit
gnatsd.varz.routes (gauge)	The number of current routes shown as unit
gnatsd.varz.slow_consumers (count)	The number of slow consumers ever connected shown as unit
gnatsd.varz.subscriptions (gauge)	The number of subscriptions currently handled shown as unit

Note: If you use custom Nats cluster names, your metrics may look like this:
gnatsd.connz.connections.cluster_name.in_msgs

Events

The gnatsd check does not include any events.

Service Checks

This gnatsd check tags all service checks it collects with:

server_name:<server_name_in_yaml>
url:<host_in_yaml>

gnatsd.can_connect: Returns CRITICAL if the Agent fails to receive a 200 from the monitoring endpoint, otherwise returns OK.

Troubleshooting

Need help? Contact Datadog support.

Mistake in the docs? Feel free to contribute!