DataDog / dogstatsd-csharp-client

A DogStatsD client for C#/.NET
datadoghq.com
Other
94 stars 62 forks source link

DogStatsD for C

Build status

A C# DogStatsD client. DogStatsD is an extension of the StatsD metric server for Datadog.

See CHANGELOG for details.

Installation

Grab the package from NuGet, or get the source from here and build it yourself.

Platforms

DogStatsD-CSharp-Client supports the following platforms:

Configuration

At start of your application, configure an instance of DogStatsdService class like this:

// The code is located under the StatsdClient namespace
using StatsdClient;

// ...

var dogstatsdConfig = new StatsdConfig
{
    StatsdServerName = "127.0.0.1",
    StatsdPort = 8125,
};

using (var service = new DogStatsdService())
{    
    if (!service.Configure(dogstatsdConfig))
        throw new InvalidOperationException("Cannot initialize DogstatsD. Set optionalExceptionHandler argument in the `Configure` method for more information.");
}

See the full list of available DogStatsD Client instantiation parameters.

Supported environment variables:

Where StatsdServerName is the hostname or address of the StatsD server, StatsdPort is the optional DogStatsD port number, and Prefix is an optional string that is prepended to all metrics.

Usage via the DogStatsdService class or the static DogStatsd class.

For usage of DogStatsD metrics, events, and Service Checks the Agent must be running and available.

Here is an example to submit different kinds of metrics with DogStatsdService.

// The code is located under the StatsdClient namespace
using StatsdClient;

// ...

var dogstatsdConfig = new StatsdConfig
{
    StatsdServerName = "127.0.0.1",
    StatsdPort = 8125,
};

using (var service = new DogStatsdService())
{
    if (!service.Configure(dogstatsdConfig))
        throw new InvalidOperationException("Cannot initialize DogstatsD. Set optionalExceptionHandler argument in the `Configure` method for more information.");

    service.Increment("example_metric.increment", tags: new[] { "environment:dev" });
    service.Decrement("example_metric.decrement", tags: new[] { "environment:dev" });
    service.Counter("example_metric.count", 2, tags: new[] { "environment:dev" });

    var random = new Random(0);

    for (int i = 0; i < 10; i++)
    {
        service.Gauge("example_metric.gauge", i, tags: new[] { "environment:dev" });
        service.Set("example_metric.set", i, tags: new[] { "environment:dev" });
        service.Histogram("example_metric.histogram", random.Next(20), tags: new[] { "environment:dev" });
        System.Threading.Thread.Sleep(random.Next(10000));
    }
}  

Here is another example to submit different kinds of metrics with DogStatsd.

// The code is located under the StatsdClient namespace
using StatsdClient;

// ...

var dogstatsdConfig = new StatsdConfig
{
    StatsdServerName = "127.0.0.1",
    StatsdPort = 8125,
};

if (!DogStatsd.Configure(dogstatsdConfig))
    throw new InvalidOperationException("Cannot initialize DogstatsD. Set optionalExceptionHandler argument in the `Configure` method for more information.");
DogStatsd.Increment("example_metric.increment", tags: new[] { "environment:dev" });
DogStatsd.Decrement("example_metric.decrement", tags: new[] { "environment:dev" });
DogStatsd.Counter("example_metric.count", 2, tags: new[] { "environment:dev" });

var random = new Random(0);

for (int i = 0; i < 10; i++)
{
    DogStatsd.Gauge("example_metric.gauge", i, tags: new[] { "environment:dev" });
    DogStatsd.Set("example_metric.set", i, tags: new[] { "environment:dev" });
    DogStatsd.Histogram("example_metric.histogram", random.Next(20), tags: new[] { "environment:dev" });
    System.Threading.Thread.Sleep(random.Next(10000));
}

DogStatsd.Dispose(); // Flush all metrics not yet sent

Metrics

After the client is created, you can start sending custom metrics to Datadog. See the dedicated Metric Submission: DogStatsD documentation to see how to submit all supported metric types to Datadog with working code examples:

Some options are suppported when submitting metrics, like applying a Sample Rate to your metrics or Tagging your metrics with your custom Tags.

Events

After the client is created, you can start sending events to your Datadog Event Stream. See the dedicated Event Submission: DogStatsD documentation to see how to submit an event to Datadog Event Stream.

Service Checks

After the client is created, you can start sending Service Checks to Datadog. See the dedicated Service Check Submission: DogStatsD documentation to see how to submit a Service Check to Datadog.

Usage via the Statsd class

Statsd has been removed in v6.0.0 because it is not thread safe and not efficient. Use DogStatsdService or DogStatsd instead:

Unix domain socket support

The version 6 (and above) of the Agent accepts packets through a Unix Socket datagram connection. Details about the advantages of using UDS over UDP are available in the Datadog DogStatsD Unix Socket documentation.

You can use unix domain socket protocol by setting StatsdServerName property to unix://YOUR_FULL_PATH, for example unix:///tmp/dsd.socket. Note that there are three / as the path of the socket is /tmp/dsd.socket.

var dogstatsdConfig = new StatsdConfig
{    
    StatsdServerName = "unix:///tmp/dsd.socket"  
};

The property StatsdMaxUnixDomainSocketPacketSize of StatsdConfig defines the maximum size of the payload. Values higher than 8196 bytes are ignored.

The feature is not supported on Windows platform. Windows has support for unix domain socket, but not for unix domain socket of type Dgram (SocketType.Dgram).

On MacOS Mojave, setting more than 2048 bytes for StatsdMaxUnixDomainSocketPacketSize is experimental.

Client side aggregation

By default, metrics for basic types (gauges, counts, sets) are aggregated before they are sent. For example, instead of sending 3 times my_metric:10|c|#tag1:value, DogStatsD client sends my_metric:30|c|#tag1:value once. For more technical details about how client-side aggregation works see #134.

Enabling client-side aggregation has the benefit of reducing the network usage and also reducing the load for DogStatsD server (Datadog Agent).

When an application sends a lot of different contexts but each context appears with a very low frequency, enabling client-side aggregation may take more memory and more CPU. A context identifies a metric name, a tag set and a metric type. The metric datadog.dogstatsd.client.aggregated_context reported by DogStatsD C# client counts the number of contexts in memory used for client-side aggregation. There is also the metric datadog.dogstatsd.client.metrics_by_type that represents the number of metrics submitted by the client before aggregation.

Configuration

The aggregation window is two seconds by default and can be changed using the FlushInterval. Note that the aggregation window on the Agent side is 10 seconds for DogStatsD metrics. For example, setting an aggregation window of 3s in the client produces a spike in your dashboard every 30s for counts metrics (as the third 10s bucket on the Agent is received 4 samples from the client).

To disable client-side aggregation set ClientSideAggregation to null.

Testing

  1. Restore packages
    dotnet restore
  2. Run the tests
    dotnet test tests/StatsdClient.Tests/

Feedback

To suggest a feature, report a bug, or general discussion, create a new issue in the Github repo.

Credits

dogstatsd-csharp-client is forked from Goncalo Pereira's original StatsD client.

Copyright (c) 2012 Goncalo Pereira and all contributors. See MIT-LICENCE.md for further details.

Thanks to Goncalo Pereira, Anthony Steele, Darrell Mozingo, Antony Denyer, and Tim Skauge for their contributions to the original client.