问题：

为发布订阅到大查询流生成此 GCP 数据流示例时出错

秦时铭

2023-03-14

我正在尝试构建以下将发布/订阅流式传输到 BigQuery 的示例：

https://github.com/GoogleCloudPlatform/DataflowTemplates/blob/master/src/main/java/com/google/cloud/teleport/templates/PubSubToBigQuery.java

代码为：

/*
 * Copyright (C) 2018 Google Inc.
 *
 * Licensed under the Apache License, Version 2.0 (the "License"); you may not
 * use this file except in compliance with the License. You may obtain a copy of
 * the License at
 *
 * http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
 * License for the specific language governing permissions and limitations under
 * the License.
 */

package com.google.cloud.teleport.templates;

import static com.google.cloud.teleport.templates.TextToBigQueryStreaming.wrapBigQueryInsertError;

import com.google.api.services.bigquery.model.TableRow;
import com.google.cloud.teleport.coders.FailsafeElementCoder;
import com.google.cloud.teleport.templates.common.BigQueryConverters.FailsafeJsonToTableRow;
import com.google.cloud.teleport.templates.common.ErrorConverters;
import com.google.cloud.teleport.templates.common.JavascriptTextTransformer.FailsafeJavascriptUdf;
import com.google.cloud.teleport.templates.common.JavascriptTextTransformer.JavascriptTextTransformerOptions;
import com.google.cloud.teleport.util.DualInputNestedValueProvider;
import com.google.cloud.teleport.util.DualInputNestedValueProvider.TranslatorInput;
import com.google.cloud.teleport.util.ResourceUtils;
import com.google.cloud.teleport.util.ValueProviderUtils;
import com.google.cloud.teleport.values.FailsafeElement;
import com.google.common.collect.ImmutableList;
import java.nio.charset.StandardCharsets;
import org.apache.beam.sdk.Pipeline;
import org.apache.beam.sdk.PipelineResult;
import org.apache.beam.sdk.coders.CoderRegistry;
import org.apache.beam.sdk.coders.StringUtf8Coder;
import org.apache.beam.sdk.io.gcp.bigquery.BigQueryIO;
import org.apache.beam.sdk.io.gcp.bigquery.BigQueryIO.Write.CreateDisposition;
import org.apache.beam.sdk.io.gcp.bigquery.BigQueryIO.Write.WriteDisposition;
import org.apache.beam.sdk.io.gcp.bigquery.BigQueryInsertError;
import org.apache.beam.sdk.io.gcp.bigquery.InsertRetryPolicy;
import org.apache.beam.sdk.io.gcp.bigquery.WriteResult;
import org.apache.beam.sdk.io.gcp.pubsub.PubsubIO;
import org.apache.beam.sdk.io.gcp.pubsub.PubsubMessage;
import org.apache.beam.sdk.io.gcp.pubsub.PubsubMessageWithAttributesCoder;
import org.apache.beam.sdk.options.Default;
import org.apache.beam.sdk.options.Description;
import org.apache.beam.sdk.options.PipelineOptions;
import org.apache.beam.sdk.options.PipelineOptionsFactory;
import org.apache.beam.sdk.options.ValueProvider;
import org.apache.beam.sdk.transforms.DoFn;
import org.apache.beam.sdk.transforms.Flatten;
import org.apache.beam.sdk.transforms.MapElements;
import org.apache.beam.sdk.transforms.PTransform;
import org.apache.beam.sdk.transforms.ParDo;
import org.apache.beam.sdk.transforms.SerializableFunction;
import org.apache.beam.sdk.values.PCollection;
import org.apache.beam.sdk.values.PCollectionList;
import org.apache.beam.sdk.values.PCollectionTuple;
import org.apache.beam.sdk.values.TupleTag;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

/**
 * The {@link PubSubToBigQuery} pipeline is a streaming pipeline which ingests data in JSON format
 * from Cloud Pub/Sub, executes a UDF, and outputs the resulting records to BigQuery. Any errors
 * which occur in the transformation of the data or execution of the UDF will be output to a
 * separate errors table in BigQuery. The errors table will be created if it does not exist prior to
 * execution. Both output and error tables are specified by the user as template parameters.
 *
 * <p><b>Pipeline Requirements</b>
 *
 * <ul>
 *   <li>The Pub/Sub topic exists.
 *   <li>The BigQuery output table exists.
 * </ul>
 *
 * <p><b>Example Usage</b>
 *
 * <pre>
 * # Set the pipeline vars
 * PROJECT_ID=PROJECT ID HERE
 * BUCKET_NAME=BUCKET NAME HERE
 * PIPELINE_FOLDER=gs://${BUCKET_NAME}/dataflow/pipelines/pubsub-to-bigquery
 * USE_SUBSCRIPTION=true or false depending on whether the pipeline should read
 *                  from a Pub/Sub Subscription or a Pub/Sub Topic.
 *
 * # Set the runner
 * RUNNER=DataflowRunner
 *
 * # Build the template
 * mvn compile exec:java \
 * -Dexec.mainClass=com.google.cloud.teleport.templates.PubSubToBigQuery \
 * -Dexec.cleanupDaemonThreads=false \
 * -Dexec.args=" \
 * --project=${PROJECT_ID} \
 * --stagingLocation=${PIPELINE_FOLDER}/staging \
 * --tempLocation=${PIPELINE_FOLDER}/temp \
 * --templateLocation=${PIPELINE_FOLDER}/template \
 * --runner=${RUNNER}
 * --useSubscription=${USE_SUBSCRIPTION}
 * "
 *
 * # Execute the template
 * JOB_NAME=pubsub-to-bigquery-$USER-`date +"%Y%m%d-%H%M%S%z"`
 *
 * # Execute a pipeline to read from a Topic.
 * gcloud dataflow jobs run ${JOB_NAME} \
 * --gcs-location=${PIPELINE_FOLDER}/template \
 * --zone=us-east1-d \
 * --parameters \
 * "inputTopic=projects/${PROJECT_ID}/topics/input-topic-name,\
 * outputTableSpec=${PROJECT_ID}:dataset-id.output-table,\
 * outputDeadletterTable=${PROJECT_ID}:dataset-id.deadletter-table"
 *
 * # Execute a pipeline to read from a Subscription.
 * gcloud dataflow jobs run ${JOB_NAME} \
 * --gcs-location=${PIPELINE_FOLDER}/template \
 * --zone=us-east1-d \
 * --parameters \
 * "inputSubscription=projects/${PROJECT_ID}/subscriptions/input-subscription-name,\
 * outputTableSpec=${PROJECT_ID}:dataset-id.output-table,\
 * outputDeadletterTable=${PROJECT_ID}:dataset-id.deadletter-table"
 * </pre>
 */
public class PubSubToBigQuery {

  /** The log to output status messages to. */
  private static final Logger LOG = LoggerFactory.getLogger(PubSubToBigQuery.class);

  /** The tag for the main output for the UDF. */
  public static final TupleTag<FailsafeElement<PubsubMessage, String>> UDF_OUT =
      new TupleTag<FailsafeElement<PubsubMessage, String>>() {};

  /** The tag for the main output of the json transformation. */
  public static final TupleTag<TableRow> TRANSFORM_OUT = new TupleTag<TableRow>() {};

  /** The tag for the dead-letter output of the udf. */
  public static final TupleTag<FailsafeElement<PubsubMessage, String>> UDF_DEADLETTER_OUT =
      new TupleTag<FailsafeElement<PubsubMessage, String>>() {};

  /** The tag for the dead-letter output of the json to table row transform. */
  public static final TupleTag<FailsafeElement<PubsubMessage, String>> TRANSFORM_DEADLETTER_OUT =
      new TupleTag<FailsafeElement<PubsubMessage, String>>() {};

  /** The default suffix for error tables if dead letter table is not specified. */
  public static final String DEFAULT_DEADLETTER_TABLE_SUFFIX = "_error_records";

  /** Pubsub message/string coder for pipeline. */
  public static final FailsafeElementCoder<PubsubMessage, String> CODER =
      FailsafeElementCoder.of(PubsubMessageWithAttributesCoder.of(), StringUtf8Coder.of());

  /** String/String Coder for FailsafeElement. */
  public static final FailsafeElementCoder<String, String> FAILSAFE_ELEMENT_CODER =
      FailsafeElementCoder.of(StringUtf8Coder.of(), StringUtf8Coder.of());

  /**
   * The {@link Options} class provides the custom execution options passed by the executor at the
   * command-line.
   */
  public interface Options extends PipelineOptions, JavascriptTextTransformerOptions {
    @Description("Table spec to write the output to")
    ValueProvider<String> getOutputTableSpec();

    void setOutputTableSpec(ValueProvider<String> value);

    @Description("Pub/Sub topic to read the input from")
    ValueProvider<String> getInputTopic();

    void setInputTopic(ValueProvider<String> value);

    @Description(
        "The Cloud Pub/Sub subscription to consume from. "
            + "The name should be in the format of "
            + "projects/<project-id>/subscriptions/<subscription-name>.")
    ValueProvider<String> getInputSubscription();

    void setInputSubscription(ValueProvider<String> value);

    @Description(
        "This determines whether the template reads from " + "a pub/sub subscription or a topic")
    @Default.Boolean(false)
    Boolean getUseSubscription();

    void setUseSubscription(Boolean value);

    @Description(
        "The dead-letter table to output to within BigQuery in <project-id>:<dataset>.<table> "
            + "format. If it doesn't exist, it will be created during pipeline execution.")
    ValueProvider<String> getOutputDeadletterTable();

    void setOutputDeadletterTable(ValueProvider<String> value);
  }

  /**
   * The main entry-point for pipeline execution. This method will start the pipeline but will not
   * wait for it's execution to finish. If blocking execution is required, use the {@link
   * PubSubToBigQuery#run(Options)} method to start the pipeline and invoke {@code
   * result.waitUntilFinish()} on the {@link PipelineResult}.
   *
   * @param args The command-line args passed by the executor.
   */
  public static void main(String[] args) {
    Options options = PipelineOptionsFactory.fromArgs(args).withValidation().as(Options.class);

    run(options);
  }

  /**
   * Runs the pipeline to completion with the specified options. This method does not wait until the
   * pipeline is finished before returning. Invoke {@code result.waitUntilFinish()} on the result
   * object to block until the pipeline is finished running if blocking programmatic execution is
   * required.
   *
   * @param options The execution options.
   * @return The pipeline result.
   */
  public static PipelineResult run(Options options) {

    Pipeline pipeline = Pipeline.create(options);

    CoderRegistry coderRegistry = pipeline.getCoderRegistry();
    coderRegistry.registerCoderForType(CODER.getEncodedTypeDescriptor(), CODER);

    /*
     * Steps:
     *  1) Read messages in from Pub/Sub
     *  2) Transform the PubsubMessages into TableRows
     *     - Transform message payload via UDF
     *     - Convert UDF result to TableRow objects
     *  3) Write successful records out to BigQuery
     *  4) Write failed records out to BigQuery
     */

    /*
     * Step #1: Read messages in from Pub/Sub
     * Either from a Subscription or Topic
     */

    PCollection<PubsubMessage> messages = null;
    if (options.getUseSubscription()) {
      messages =
          pipeline.apply(
              "ReadPubSubSubscription",
              PubsubIO.readMessagesWithAttributes()
                  .fromSubscription(options.getInputSubscription()));
    } else {
      messages =
          pipeline.apply(
              "ReadPubSubTopic",
              PubsubIO.readMessagesWithAttributes().fromTopic(options.getInputTopic()));
    }

    PCollectionTuple convertedTableRows =
        messages
            /*
             * Step #2: Transform the PubsubMessages into TableRows
             */
            .apply("ConvertMessageToTableRow", new PubsubMessageToTableRow(options));

    /*
     * Step #3: Write the successful records out to BigQuery
     */
    WriteResult writeResult =
        convertedTableRows
            .get(TRANSFORM_OUT)
            .apply(
                "WriteSuccessfulRecords",
                BigQueryIO.writeTableRows()
                    .withoutValidation()
                    .withCreateDisposition(CreateDisposition.CREATE_NEVER)
                    .withWriteDisposition(WriteDisposition.WRITE_APPEND)
                    .withExtendedErrorInfo()
                    .withMethod(BigQueryIO.Write.Method.STREAMING_INSERTS)
                    .withFailedInsertRetryPolicy(InsertRetryPolicy.retryTransientErrors())
                    .to(options.getOutputTableSpec()));

    /*
     * Step 3 Contd.
     * Elements that failed inserts into BigQuery are extracted and converted to FailsafeElement
     */
    PCollection<FailsafeElement<String, String>> failedInserts =
        writeResult
            .getFailedInsertsWithErr()
            .apply(
                "WrapInsertionErrors",
                MapElements.into(FAILSAFE_ELEMENT_CODER.getEncodedTypeDescriptor())
                    .via((BigQueryInsertError e) -> wrapBigQueryInsertError(e)))
            .setCoder(FAILSAFE_ELEMENT_CODER);

    /*
     * Step #4: Write records that failed table row transformation
     * or conversion out to BigQuery deadletter table.
     */
    PCollectionList.of(
            ImmutableList.of(
                convertedTableRows.get(UDF_DEADLETTER_OUT),
                convertedTableRows.get(TRANSFORM_DEADLETTER_OUT)))
        .apply("Flatten", Flatten.pCollections())
        .apply(
            "WriteFailedRecords",
            ErrorConverters.WritePubsubMessageErrors.newBuilder()
                .setErrorRecordsTable(
                    ValueProviderUtils.maybeUseDefaultDeadletterTable(
                        options.getOutputDeadletterTable(),
                        options.getOutputTableSpec(),
                        DEFAULT_DEADLETTER_TABLE_SUFFIX))
                .setErrorRecordsTableSchema(ResourceUtils.getDeadletterTableSchemaJson())
                .build());

    // 5) Insert records that failed insert into deadletter table
    failedInserts.apply(
        "WriteFailedRecords",
        ErrorConverters.WriteStringMessageErrors.newBuilder()
            .setErrorRecordsTable(
                ValueProviderUtils.maybeUseDefaultDeadletterTable(
                    options.getOutputDeadletterTable(),
                    options.getOutputTableSpec(),
                    DEFAULT_DEADLETTER_TABLE_SUFFIX))
            .setErrorRecordsTableSchema(ResourceUtils.getDeadletterTableSchemaJson())
            .build());

    return pipeline.run();
  }

  /**
   * If deadletterTable is available, it is returned as is, otherwise outputTableSpec +
   * defaultDeadLetterTableSuffix is returned instead.
   */
  private static ValueProvider<String> maybeUseDefaultDeadletterTable(
      ValueProvider<String> deadletterTable,
      ValueProvider<String> outputTableSpec,
      String defaultDeadLetterTableSuffix) {
    return DualInputNestedValueProvider.of(
        deadletterTable,
        outputTableSpec,
        new SerializableFunction<TranslatorInput<String, String>, String>() {
          @Override
          public String apply(TranslatorInput<String, String> input) {
            String userProvidedTable = input.getX();
            String outputTableSpec = input.getY();
            if (userProvidedTable == null) {
              return outputTableSpec + defaultDeadLetterTableSuffix;
            }
            return userProvidedTable;
          }
        });
  }

  /**
   * The {@link PubsubMessageToTableRow} class is a {@link PTransform} which transforms incoming
   * {@link PubsubMessage} objects into {@link TableRow} objects for insertion into BigQuery while
   * applying an optional UDF to the input. The executions of the UDF and transformation to {@link
   * TableRow} objects is done in a fail-safe way by wrapping the element with it's original payload
   * inside the {@link FailsafeElement} class. The {@link PubsubMessageToTableRow} transform will
   * output a {@link PCollectionTuple} which contains all output and dead-letter {@link
   * PCollection}.
   *
   * <p>The {@link PCollectionTuple} output will contain the following {@link PCollection}:
   *
   * <ul>
   *   <li>{@link PubSubToBigQuery#UDF_OUT} - Contains all {@link FailsafeElement} records
   *       successfully processed by the optional UDF.
   *   <li>{@link PubSubToBigQuery#UDF_DEADLETTER_OUT} - Contains all {@link FailsafeElement}
   *       records which failed processing during the UDF execution.
   *   <li>{@link PubSubToBigQuery#TRANSFORM_OUT} - Contains all records successfully converted from
   *       JSON to {@link TableRow} objects.
   *   <li>{@link PubSubToBigQuery#TRANSFORM_DEADLETTER_OUT} - Contains all {@link FailsafeElement}
   *       records which couldn't be converted to table rows.
   * </ul>
   */
  static class PubsubMessageToTableRow
      extends PTransform<PCollection<PubsubMessage>, PCollectionTuple> {

    private final Options options;

    PubsubMessageToTableRow(Options options) {
      this.options = options;
    }

    @Override
    public PCollectionTuple expand(PCollection<PubsubMessage> input) {

      PCollectionTuple udfOut =
          input
              // Map the incoming messages into FailsafeElements so we can recover from failures
              // across multiple transforms.
              .apply("MapToRecord", ParDo.of(new PubsubMessageToFailsafeElementFn()))
              .apply(
                  "InvokeUDF",
                  FailsafeJavascriptUdf.<PubsubMessage>newBuilder()
                      .setFileSystemPath(options.getJavascriptTextTransformGcsPath())
                      .setFunctionName(options.getJavascriptTextTransformFunctionName())
                      .setSuccessTag(UDF_OUT)
                      .setFailureTag(UDF_DEADLETTER_OUT)
                      .build());

      // Convert the records which were successfully processed by the UDF into TableRow objects.
      PCollectionTuple jsonToTableRowOut =
          udfOut
              .get(UDF_OUT)
              .apply(
                  "JsonToTableRow",
                  FailsafeJsonToTableRow.<PubsubMessage>newBuilder()
                      .setSuccessTag(TRANSFORM_OUT)
                      .setFailureTag(TRANSFORM_DEADLETTER_OUT)
                      .build());

      // Re-wrap the PCollections so we can return a single PCollectionTuple
      return PCollectionTuple.of(UDF_OUT, udfOut.get(UDF_OUT))
          .and(UDF_DEADLETTER_OUT, udfOut.get(UDF_DEADLETTER_OUT))
          .and(TRANSFORM_OUT, jsonToTableRowOut.get(TRANSFORM_OUT))
          .and(TRANSFORM_DEADLETTER_OUT, jsonToTableRowOut.get(TRANSFORM_DEADLETTER_OUT));
    }
  }

  /**
   * The {@link PubsubMessageToFailsafeElementFn} wraps an incoming {@link PubsubMessage} with the
   * {@link FailsafeElement} class so errors can be recovered from and the original message can be
   * output to a error records table.
   */
  static class PubsubMessageToFailsafeElementFn
      extends DoFn<PubsubMessage, FailsafeElement<PubsubMessage, String>> {
    @ProcessElement
    public void processElement(ProcessContext context) {
      PubsubMessage message = context.element();
      context.output(
          FailsafeElement.of(message, new String(message.getPayload(), StandardCharsets.UTF_8)));
    }
  }
}

使用以下maven

<?xml version="1.0" encoding="UTF-8"?>

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
  <modelVersion>4.0.0</modelVersion>

  <groupId>pubsub_bigquery_dataflow</groupId>
  <artifactId>pubsub_bigquery_dataflow</artifactId>
  <version>1.0-SNAPSHOT</version>

  <name>pubsub_bigquery_dataflow</name>
  <!-- FIXME change it to the project's website -->
  <url>http://www.example.com</url>

  <properties>
    <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
    <maven.compiler.source>1.7</maven.compiler.source>
    <maven.compiler.target>1.7</maven.compiler.target>
  </properties>

  <dependencies>
    <dependency>
      <groupId>junit</groupId>
      <artifactId>junit</artifactId>
      <version>4.11</version>
      <scope>test</scope>
    </dependency>
    <dependency>
      <groupId>com.google.cloud.dataflow</groupId>
      <artifactId>google-cloud-dataflow-java-sdk-all</artifactId>
      <version>2.5.0</version>
    </dependency>
  </dependencies>

  <build>
    <pluginManagement><!-- lock down plugins versions to avoid using Maven defaults (may be moved to parent pom) -->
      <plugins>
        <!-- clean lifecycle, see https://maven.apache.org/ref/current/maven-core/lifecycles.html#clean_Lifecycle -->
        <plugin>
          <artifactId>maven-clean-plugin</artifactId>
          <version>3.1.0</version>
        </plugin>
        <!-- default lifecycle, jar packaging: see https://maven.apache.org/ref/current/maven-core/default-bindings.html#Plugin_bindings_for_jar_packaging -->
        <plugin>
          <artifactId>maven-resources-plugin</artifactId>
          <version>3.0.2</version>
        </plugin>
        <plugin>
          <artifactId>maven-compiler-plugin</artifactId>
          <version>3.8.0</version>
        </plugin>
        <plugin>
          <artifactId>maven-surefire-plugin</artifactId>
          <version>2.22.1</version>
        </plugin>
        <plugin>
          <artifactId>maven-jar-plugin</artifactId>
          <version>3.0.2</version>
        </plugin>
        <plugin>
          <artifactId>maven-install-plugin</artifactId>
          <version>2.5.2</version>
        </plugin>
        <plugin>
          <artifactId>maven-deploy-plugin</artifactId>
          <version>2.8.2</version>
        </plugin>
        <!-- site lifecycle, see https://maven.apache.org/ref/current/maven-core/lifecycles.html#site_Lifecycle -->
        <plugin>
          <artifactId>maven-site-plugin</artifactId>
          <version>3.7.1</version>
        </plugin>
        <plugin>
          <artifactId>maven-project-info-reports-plugin</artifactId>
          <version>3.0.0</version>
        </plugin>
      </plugins>
    </pluginManagement>
  </build>
</project>

但是我得到错误，比如“com . Google . cloud . teleport . coders”包不存在。该示例也出现在https://cloud . Google . com/data flow/docs/guides/templates/provided-templates # cloudpubsubtobigquery，但没有关于需要哪些jar或任何依赖文件的说明。

共有2个答案

公冶兴文

2023-03-14

检查pom。DataflowTemplates Github中的xml似乎缺少一些依赖项。

弓方伟

2023-03-14

您尝试构建的类不能仅使用该文件构建。它引用了出现在存储库本身中的其他几个类，例如，com.google.cloud.teleport.coders。存储库的主要说明说，必须使用命令mvn干净编译构建整个项目。然后，这些说明提供了构建和暂存模板文件所需的命令。如果要单独拉出其中一个模板，则需要在 pom 文件中包含外部依赖项，并提取它们生成的本地依赖项。导入语句应指示依赖关系。那些在微软网中的人都在同一个存储库中。其余的将在主pom.xml中引用。

为发布订阅到大查询流生成此 GCP 数据流示例时出错

共有2个答案

相关问答

相关文章

相关阅读

相关工具

相关文档