VMware has ended active development of this project, this repository will no longer be updated.

wavefront-grpc-sdk-java

The Wavefront gRPC SDK for Java is a library that collects out-of-the-box metrics, histograms, and trace data from gRPC operations in your Java application, and reports that data to Wavefront. You can analyze the telemetry data in Wavefront to better understand how your application is performing in production.

Maven

If you are using Maven, add following maven dependency to your pom.xml.

<dependency>
    <groupId>com.wavefront</groupId>
    <artifactId>wavefront-grpc-sdk-java</artifactId>
    <version>$releaseVersion</version>
</dependency>

Replace $releaseVersion with the latest version available on maven.

Setup Steps

Follow the steps below to set up objects for collecting metrics, histograms, and trace data from gRPC requests and responses in a microservice.

For each gRPC-based microservice, add the dependency if you have not already done so, and then perform the following steps:

1. Create an ApplicationTags instance, which specifies metadata about your application.
2. Create a WavefrontSender for sending data to Wavefront.
3. Create a WavefrontGrpcReporter for reporting gRPC metrics and histograms to Wavefront.
4. Create a WavefrontClientInterceptor or a WavefrontServerTracerFactory, as appropriate:

   • To instrument a gRPC client, create a WavefrontClientInterceptor to collect telemetry data from sent requests and received responses.

   • To instrument a gRPC server, create a WavefrontServerTracerFactory to collect telemetry data from the received requests and sent responses.

For the details of each step, see the sections below.

1. Set Up Application Tags

Application tags determine the metadata (point tags and span tags) that are included with every metric/histogram/span reported to Wavefront. These tags enable you to filter and query the reported data in Wavefront.

You encapsulate application tags in an ApplicationTags object. See Instantiating ApplicationTags for details.

2. Set Up a WavefrontSender

A WavefrontSender object implements the low-level interface for sending data to Wavefront. You can choose to send data to Wavefront using either the Wavefront proxy or direct ingestion.

• If you have already set up a WavefrontSender for another SDK that will run in the same JVM, use that one. (For details about sharing a WavefrontSender instance, see Share a WavefrontSender.)

• Otherwise, follow the steps in Set Up a WavefrontSender.

3. Set Up a WavefrontGrpcReporter

A WavefrontGrpcReporter reports trace data to Wavefront. To build a WavefrontGrpcReporter, you specify:

• An ApplicationTags object (see above)
• A WavefrontSender object (see above).

You can optionally specify:

• A nondefault source for the reported data. If you omit the source, the host name is automatically used.
• A nondefault reporting interval, which controls how often data is reported to the WavefrontSender. The reporting interval determines the timestamps on the data sent to Wavefront. If you omit the reporting interval, data is reported once a minute.

ApplicationTags applicationTags = buildTags(); // pseudocode; see above

// Create WavefrontGrpcReporter.Builder using applicationTags.
WavefrontGrpcReporter.Builder wfGrpcReporterBuilder = new WavefrontGrpcReporter.Builder(applicationTags);

// Optionally set a nondefault source name for your metrics and histograms. Omit this statement to use the host name.
wfGrpcReporterBuilder.withSource("mySource");

// Optionally change the reporting interval to 30 seconds. Default is 1 minute.
wfGrpcReporterBuilder.reportingIntervalSeconds(30);

// Create a WavefrontGrpcReporter with a WavefrontSender
WavefrontGrpcReporter wfGrpcReporter = wfGrpcReporterBuilder.build(wavefrontSender);

4. Set Up Client or Server Instrumentation

Choose one or both of the following options based on your microservice's behavior.

Option 1. Instrument a gRPC Client

If you are instrumenting a gRPC client, set up a WavefrontClientInterceptor to intercept the gRPC channel and collect telemetry data from all requests that the client sends. Specify true or false to enable or disable streaming specific stats.

// Create a WavefrontClientInterceptor using ApplicationTags and a WavefrontGrpcReporter.
WavefrontClientInterceptor wfClientInterceptor = new WavefrontServerTracerFactory(wfGrpcReporter, applicationTags, true);

Option 2. Instrument a gRPC Server

If you are instrumenting a gRPC server, set up a WavefrontServerTracerFactory and to collect telemetry data from all requests that the client receives. Specify true or false to enable or disable streaming specific stats.

// Create a WavefrontServerTracerFactory using the application tags and a WavefrontGrpcReporter.
WavefrontServerTracerFactory wfServerTracerFactory = new WavefrontServerTracerFactory(wfGrpcReporter, applicationTags, true);

Metrics and Histograms Sent From gRPC Operations

Let's say you have an order-managing gRPC service with the following proto schema:

syntax = "proto3";

option java_package = "com.wf.examples.inventory"

package com.ordering

// The order service definition.
service OrderManager {
   // Get single order status
   rpc getOrder (OrderDetails) returns (Order) {}

  // Get a stream of orders, which is a server streaming RPC
  rpc getAllOrders (OrderFilter) returns (stream Order) {}
}

When instrumented using the Wavefront gRPC SDK for Java, the gRPC service sends Server Metrics or Client Metrics as appropriate.

Note: For gRPC stats, grpc.service is the service name defined in the proto schema and service refers to user-provided value in ApplicationTags.

Server Metrics

Assume this gRPC Service server is:

1. Part of the Ordering application
2. Running inside the Inventory microservice
3. Deployed in us-west-1 cluster
4. Serviced by primary shard
5. On source = host-1
6. And the API call returns OK gRPC status code.

If this gRPC service is instrumented as a server using the Wavefront gRPC SDK for Java, the stats listed below are sent to Wavefront.

Request Gauges

Granular Response Related Metrics

Granular Response Latency Histogram

Granular Request and Response Payload Size Histograms

Granular Streaming Metrics and Histograms

Overall Response Metrics

Overall Error Metrics

Client Metrics

Assume this gRPC Service client accessing the OrderManager is:

1. Part of the Ordering application
2. Running inside the Billing microservice
3. Deployed in us-west-1 cluster
4. Serviced by primary shard
5. On source = host-1
6. And the API call returns OK gRPC status code.

If this gRPC service is instrumented as a client using the Wavefront gRPC SDK for Java, the stats listed below are sent to Wavefront.

Request Gauges

Granular Response Related Metrics

Granular Response Latency Histogram

Granular Request and Response Payload Size Histograms

Granular Streaming Metrics and Histograms

Overall Response Metrics

Overall Error Metrics