Compress messages

If you are using Pub/Sub to publish messages that amount to a lot of data, you can use gRPC to compress your data to save networking costs before your publisher client sends out the publish request. The Pub/Sub compression for gRPC uses the Gzip algorithm.

This document provides information about compressing messages published to a topic.

About compressing messages

The compression ratio for using the gRPC client-side compression feature is different for different publisher clients and dependent on the following factors:

  • Amount of data. The compression ratio improves when the size of the payload increases from a few hundred bytes to many kilobytes of data. The batch settings of a publish request decides the amount of data that is included in each publish request. We recommend that you turn on batch settings in conjunction with gRPC compression to get the best results.

  • Type of data. Text-based data such as JSON or XML are more compressible compared to binary data such as images.

If your publisher client is on Google Cloud, you can use the Sent bytes (instance/network/sent_bytes_count) metric to measure the publishing throughput in bytes. If your publisher client is on a different application, you must use the client-specific tools for making the measurement.

The code sample in this section shows a sample Java client library code snippet that also includes gRPC compression.

Before you begin

Before configuring the publish workflow, ensure you have completed the following tasks:

Required roles

To get the permissions that you need to compress messages, ask your administrator to grant you the Pub/Sub Publisher (roles/pubsub.publisher) IAM role on topic. For more information about granting roles, see Manage access.

You might also be able to get the required permissions through custom roles or other predefined roles.

You need additional permissions to create or update topics and subscriptions.

Compress a message

C++

Before trying this sample, follow the C++ setup instructions in Quickstart: Using Client Libraries. For more information, see the Pub/Sub C++ API reference documentation. 

namespace g = ::google::cloud;
namespace pubsub = ::google::cloud::pubsub;
[](std::string project_id, std::string topic_id) {
  auto topic = pubsub::Topic(std::move(project_id), std::move(topic_id));
  auto publisher = pubsub::Publisher(pubsub::MakePublisherConnection(
      std::move(topic),
      g::Options{}
          // Compress any batch of messages over 10 bytes. By default, no
          // messages are compressed, set this to 0 to compress all batches,
          // regardless of their size.
          .set<pubsub::CompressionThresholdOption>(10)
          // Compress using the GZIP algorithm. By default, the library uses
          // GRPC_COMPRESS_DEFLATE.
          .set<pubsub::CompressionAlgorithmOption>(GRPC_COMPRESS_GZIP)));
  auto message_id = publisher.Publish(
      pubsub::MessageBuilder{}.SetData("Hello World!").Build());
  auto done = message_id.then([](g::future<g::StatusOr<std::string>> f) {
    auto id = f.get();
    if (!id) throw std::move(id).status();
    std::cout << "Hello World! published with id=" << *id << "\n";
  });
  // Block until the message is published
  done.get();
}

Java

Before trying this sample, follow the Java setup instructions in Quickstart: Using Client Libraries. For more information, see the Pub/Sub Java API reference documentation. 

import com.google.api.core.ApiFuture;
import com.google.cloud.pubsub.v1.Publisher;
import com.google.protobuf.ByteString;
import com.google.pubsub.v1.PubsubMessage;
import com.google.pubsub.v1.TopicName;
import java.io.IOException;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.TimeUnit;

public class PublishWithGrpcCompressionExample {
  public static void main(String... args) throws Exception {
    // TODO(developer): Replace these variables before running the sample.
    String projectId = "your-project-id";
    // Choose an existing topic.
    String topicId = "your-topic-id";

    publishWithGrpcCompressionExample(projectId, topicId);
  }

  public static void publishWithGrpcCompressionExample(String projectId, String topicId)
      throws IOException, ExecutionException, InterruptedException {
    TopicName topicName = TopicName.of(projectId, topicId);

    // Create a publisher and set enable compression to true.
    Publisher publisher = null;
    try {
      // Enable compression and configure the compression threshold to 10 bytes (default to 240 B).
      // Publish requests of sizes > 10 B (excluding the request headers) will get compressed.
      // The number of messages in a publish request is determined by publisher batch settings.
      // Batching is turned off by default, i.e. each publish request contains only one message.
      publisher =
          Publisher.newBuilder(topicName)
              .setEnableCompression(true)
              .setCompressionBytesThreshold(10L)
              .build();

      byte[] bytes = new byte[1024];
      ByteString data = ByteString.copyFrom(bytes);
      PubsubMessage pubsubMessage = PubsubMessage.newBuilder().setData(data).build();

      // Once published, returns a server-assigned message id (unique within the topic).
      // You can look up the actual size of the outbound data using the Java Logging API.
      // Configure logging properties as shown in
      // https://github.com/googleapis/java-pubsub/tree/main/samples/snippets/src/main/resources/logging.properties
      // and look for "OUTBOUND DATA" with "length=" in the output log.
      ApiFuture<String> messageIdFuture = publisher.publish(pubsubMessage);
      String messageId = messageIdFuture.get();
      System.out.println("Published a compressed message of message ID: " + messageId);
    } finally {
      if (publisher != null) {
        // When finished with the publisher, shutdown to free up resources.
        publisher.shutdown();
        publisher.awaitTermination(1, TimeUnit.MINUTES);
      }
    }
  }
}

What's next

To learn how to configure advanced publishing options, see the following: