Vertex AI Feature Store Bigtable Gateway

Custom Java library for accessing directly online store(Bigtable).

By Vertex AI Feature Store

Getting Started

This feature is disabled by default. To enable it for your project, please contact vertex-feature-store-feedback@google.com.

Add dependency

If you're using Maven, add the following to your dependencies:

<dependency>
  <groupId>com.google.cloud</groupId>
  <artifactId>google-cloud-aiplatform-fs</artifactId>
  <version>0.1.0</version>
  <exclusions>
    <exclusion>
      <groupId>com.google.api.grpc</groupId>
      <artifactId>proto-google-cloud-aiplatform-v1</artifactId>
    </exclusion>
    <exclusion>
      <groupId>com.google.api.grpc</groupId>
      <artifactId>grpc-google-cloud-aiplatform-v1</artifactId>
    </exclusion>
    <exclusion>
      <groupId>com.google.cloud</groupId>
      <artifactId>gapic-google-cloud-aiplatform-v1</artifactId>
    </exclusion>
  </exclusions>
</dependency>

Create a client and call FetchFeatureValues APIs

import com.google.cloud.aiplatform.fs.FeatureOnlineStoreDirectClient;
import com.google.cloud.aiplatform.v1.FeatureValue;
import com.google.cloud.aiplatform.v1.FeatureView;
import com.google.cloud.aiplatform.v1.FeatureViewDataFormat;
import com.google.cloud.aiplatform.v1.FeatureViewDataKey;
import com.google.cloud.aiplatform.v1.FetchFeatureValuesRequest;
import com.google.cloud.aiplatform.v1.FetchFeatureValuesResponse;

public class DirectReadExample {

  private static final String FEATURE_VIEW_NAME =
      "projects/<PROJECT>/locations/<LOCATION>/featureOnlineStores/<FOS>/featureViews/<FV>";
  private FeatureOnlineStoreDirectClient client;

  DirectReadExample() {
    try {
      client = FeatureOnlineStoreDirectClient.create(FEATURE_VIEW_NAME);
    } catch (Exception e) {
      System.out.println("Could not create client" + e);
    }
  }

  public void fetchFeatureValuesSync(String key) {
    FetchFeatureValuesRequest request =
        FetchFeatureValuesRequest.newBuilder()
            .setFeatureView(FEATURE_VIEW_NAME)
            .setDataFormat(FeatureViewDataFormat.KEY_VALUE)
            .setDataKey(FeatureViewDataKey.newBuilder().setKey(key).build())
            .build();
    try {
      FetchFeatureValuesResponse response = client.fetchFeatureValues(request);
      System.out.println(response.toString());
    } catch (Exception e) {
      System.out.println("Error fetching features");
    }
  }

  public static void main(String[] args) {
    DirectReadExample ex = new DirectReadExample();
    String key = "your data key to read from FeatureView";
    ex.fetchFeatureValuesSync(key);
    System.out.println("Test done");
  }
}

Configure Bigtable connection

Below code is an example for configuring Bigtable connection.

import com.google.api.gax.grpc.ChannelPoolSettings;
import com.google.api.gax.grpc.InstantiatingGrpcChannelProvider;
import com.google.api.gax.retrying.RetrySettings;
import com.google.api.gax.rpc.StatusCode;
import com.google.api.gax.rpc.StatusCode.Code;
import com.google.api.gax.rpc.TransportChannelProvider;
import com.google.cloud.aiplatform.fs.DirectClientSettings;
import com.google.cloud.aiplatform.fs.FeatureOnlineStoreDirectClient;

public class DirectReadExample {

  private static final String FEATURE_VIEW_NAME =
      "projects/<PROJECT>/locations/<LOCATION>/featureOnlineStores/<FOS>/featureViews/<FV>";
  private FeatureOnlineStoreDirectClient client;

  DirectReadExample() {
    // Add configuration for the channel pool.
    ChannelPoolSettings channelPoolSettings =
        ChannelPoolSettings.builder()
            .setInitialChannelCount(3)
            .setMaxChannelCount(10)
            .setMinChannelCount(3)
            .build();
    // Add configuration for the retry behavior.
    RetrySettings retrySettings =
        RetrySettings.newBuilder()
            .setInitialRetryDelayDuration(Duration.ofSeconds(1))
            .setMaxAttempts(10)
            .setMaxRetryDelayDuration(Duration.ofSeconds(5))
            .build();

    DirectClientSettings settings =
        new DirectClientSettings.Builder()
            .setChannelPoolSettings(channelPoolSettings)
            .setRetrySettings(retrySettings)
            // Configure which error codes to enable retries.
            .setRetryableCodes(Code.ABORTED, Code.INTERNAL, Code.CANCELLED)
            .build();

    try {
      client = FeatureOnlineStoreDirectClient.create(FEATURE_VIEW_NAME, settings);
    } catch (Exception e) {
      System.out.println("Could not create client" + e);
    }
  }
}

Contribute to this library

This Vertex AI Feature Store Java SDK will not accept contributions.

License

Apache 2.0 - See [LICENSE][license] for more information.