All Downloads are FREE. Search and download functionalities are using the official Maven repository.

com.google.cloud.bigquery.ConnectionImpl Maven / Gradle / Ivy

There is a newer version: 2.44.0
Show newest version
/*
 * Copyright 2021 Google LLC
 *
 * 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.bigquery;

import static com.google.cloud.RetryHelper.runWithRetries;
import static java.net.HttpURLConnection.HTTP_NOT_FOUND;

import com.google.api.core.BetaApi;
import com.google.api.core.InternalApi;
import com.google.api.gax.core.FixedCredentialsProvider;
import com.google.api.services.bigquery.model.GetQueryResultsResponse;
import com.google.api.services.bigquery.model.JobConfigurationQuery;
import com.google.api.services.bigquery.model.QueryParameter;
import com.google.api.services.bigquery.model.QueryRequest;
import com.google.api.services.bigquery.model.TableDataList;
import com.google.api.services.bigquery.model.TableRow;
import com.google.cloud.RetryHelper;
import com.google.cloud.Tuple;
import com.google.cloud.bigquery.JobStatistics.QueryStatistics;
import com.google.cloud.bigquery.JobStatistics.SessionInfo;
import com.google.cloud.bigquery.spi.v2.BigQueryRpc;
import com.google.cloud.bigquery.storage.v1.ArrowRecordBatch;
import com.google.cloud.bigquery.storage.v1.ArrowSchema;
import com.google.cloud.bigquery.storage.v1.BigQueryReadClient;
import com.google.cloud.bigquery.storage.v1.BigQueryReadSettings;
import com.google.cloud.bigquery.storage.v1.CreateReadSessionRequest;
import com.google.cloud.bigquery.storage.v1.DataFormat;
import com.google.cloud.bigquery.storage.v1.ReadRowsRequest;
import com.google.cloud.bigquery.storage.v1.ReadRowsResponse;
import com.google.cloud.bigquery.storage.v1.ReadSession;
import com.google.common.annotations.VisibleForTesting;
import com.google.common.base.Function;
import com.google.common.base.Strings;
import com.google.common.collect.ImmutableList;
import com.google.common.collect.Iterables;
import com.google.common.collect.Lists;
import com.google.common.collect.Maps;
import com.google.common.util.concurrent.FutureCallback;
import com.google.common.util.concurrent.Futures;
import com.google.common.util.concurrent.ListenableFuture;
import com.google.common.util.concurrent.ListeningExecutorService;
import com.google.common.util.concurrent.MoreExecutors;
import java.io.IOException;
import java.math.BigInteger;
import java.util.AbstractList;
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashMap;
import java.util.LinkedList;
import java.util.List;
import java.util.Map;
import java.util.Queue;
import java.util.UUID;
import java.util.concurrent.BlockingQueue;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.LinkedBlockingDeque;
import java.util.concurrent.TimeUnit;
import java.util.logging.Level;
import java.util.logging.Logger;
import java.util.stream.Collectors;
import org.apache.arrow.memory.BufferAllocator;
import org.apache.arrow.memory.RootAllocator;
import org.apache.arrow.vector.FieldVector;
import org.apache.arrow.vector.VectorLoader;
import org.apache.arrow.vector.VectorSchemaRoot;
import org.apache.arrow.vector.ipc.ReadChannel;
import org.apache.arrow.vector.ipc.message.MessageSerializer;
import org.apache.arrow.vector.types.pojo.Field;
import org.apache.arrow.vector.util.ByteArrayReadableSeekableByteChannel;

/** Implementation for {@link Connection}, the generic BigQuery connection API (not JDBC). */
class ConnectionImpl implements Connection {

  private final ConnectionSettings connectionSettings;
  private final BigQueryOptions bigQueryOptions;
  private final BigQueryRpc bigQueryRpc;
  private final BigQueryRetryConfig retryConfig;
  private final int bufferSize; // buffer size in Producer Thread
  private final int MAX_PROCESS_QUERY_THREADS_CNT = 5;
  private final ExecutorService queryTaskExecutor =
      Executors.newFixedThreadPool(MAX_PROCESS_QUERY_THREADS_CNT);
  private final Logger logger = Logger.getLogger(this.getClass().getName());
  private BigQueryReadClient bqReadClient;
  private static final long EXECUTOR_TIMEOUT_SEC = 10;
  private BlockingQueue>
      bufferFvl; // initialized lazily iff we end up using the tabledata.list end point
  private BlockingQueue
      bufferRow; // initialized lazily iff we end up using Read API

  ConnectionImpl(
      ConnectionSettings connectionSettings,
      BigQueryOptions bigQueryOptions,
      BigQueryRpc bigQueryRpc,
      BigQueryRetryConfig retryConfig) {
    this.connectionSettings = connectionSettings;
    this.bigQueryOptions = bigQueryOptions;
    this.bigQueryRpc = bigQueryRpc;
    this.retryConfig = retryConfig;
    // Sets a reasonable buffer size (a blocking queue) if user input is suboptimal
    this.bufferSize =
        (connectionSettings == null
                || connectionSettings.getNumBufferedRows() == null
                || connectionSettings.getNumBufferedRows() < 10000
            ? 20000
            : Math.min(connectionSettings.getNumBufferedRows() * 2, 100000));
  }

  /**
   * This method returns the number of records to be stored in the buffer and it ensures that it is
   * between a reasonable range
   *
   * @return The max number of records to be stored in the buffer
   */
  private int getBufferSize() {
    return (connectionSettings == null
            || connectionSettings.getNumBufferedRows() == null
            || connectionSettings.getNumBufferedRows() < 10000
        ? 20000
        : Math.min(connectionSettings.getNumBufferedRows() * 2, 100000));
  }
  /**
   * Cancel method shutdowns the pageFetcher and producerWorker threads gracefully using interrupt.
   * The pageFetcher threat will not request for any subsequent threads after interrupting and
   * shutdown as soon as any ongoing RPC call returns. The producerWorker will not populate the
   * buffer with any further records and clear the buffer, put a EoF marker and shutdown.
   *
   * @return Boolean value true if the threads were interrupted
   * @throws BigQuerySQLException
   */
  @BetaApi
  @Override
  public synchronized boolean close() throws BigQuerySQLException {
    flagEndOfStream(); // an End of Stream flag in the buffer so that the `ResultSet.next()` stops
    // advancing the cursor
    queryTaskExecutor.shutdownNow();
    try {
      if (queryTaskExecutor.awaitTermination(EXECUTOR_TIMEOUT_SEC, TimeUnit.SECONDS)) {
        return true;
      } // else queryTaskExecutor.isShutdown() will be returned outside this try block
    } catch (InterruptedException e) {
      logger.log(
          Level.WARNING,
          "\n" + Thread.currentThread().getName() + " Exception while awaitTermination",
          e); // Logging InterruptedException instead of throwing the exception back, close method
      // will return queryTaskExecutor.isShutdown()
    }
    return queryTaskExecutor.isShutdown(); // check if the executor has been shutdown
  }

  /**
   * This method runs a dry run query
   *
   * @param sql SQL SELECT statement
   * @return BigQueryDryRunResult containing List and Schema
   * @throws BigQuerySQLException
   */
  @BetaApi
  @Override
  public BigQueryDryRunResult dryRun(String sql) throws BigQuerySQLException {
    com.google.api.services.bigquery.model.Job dryRunJob = createDryRunJob(sql);
    Schema schema = Schema.fromPb(dryRunJob.getStatistics().getQuery().getSchema());
    List queryParametersPb =
        dryRunJob.getStatistics().getQuery().getUndeclaredQueryParameters();
    List queryParameters =
        queryParametersPb == null
            ? Collections.emptyList()
            : Lists.transform(queryParametersPb, QUERY_PARAMETER_FROM_PB_FUNCTION);
    QueryStatistics queryStatistics = JobStatistics.fromPb(dryRunJob);
    SessionInfo sessionInfo =
        queryStatistics.getSessionInfo() == null ? null : queryStatistics.getSessionInfo();
    BigQueryResultStats bigQueryResultStats =
        new BigQueryResultStatsImpl(queryStatistics, sessionInfo);
    return new BigQueryDryRunResultImpl(schema, queryParameters, bigQueryResultStats);
  }

  /**
   * This method executes a SQL SELECT query
   *
   * @param sql SQL SELECT statement
   * @return BigQueryResult containing the output of the query
   * @throws BigQuerySQLException
   */
  @BetaApi
  @Override
  public BigQueryResult executeSelect(String sql) throws BigQuerySQLException {
    return getExecuteSelectResponse(sql, null, null);
  }

  /**
   * This method executes a SQL SELECT query
   *
   * @param sql SQL SELECT query
   * @param parameters named or positional parameters. The set of query parameters must either be
   *     all positional or all named parameters.
   * @param labels the labels associated with this query. You can use these to organize and group
   *     your query jobs. Label keys and values can be no longer than 63 characters, can only
   *     contain lowercase letters, numeric characters, underscores and dashes. International
   *     characters are allowed. Label values are optional and Label is a Varargs. You should pass
   *     all the Labels in a single Map .Label keys must start with a letter and each label in the
   *     list must have a different key.
   * @return BigQueryResult containing the output of the query
   * @throws BigQuerySQLException
   */
  @BetaApi
  @Override
  public BigQueryResult executeSelect(
      String sql, List parameters, Map... labels)
      throws BigQuerySQLException {
    return getExecuteSelectResponse(sql, parameters, labels);
  }

  private BigQueryResult getExecuteSelectResponse(
      String sql, List parameters, Map... labels)
      throws BigQuerySQLException {
    Map labelMap = null;
    if (labels != null
        && labels.length == 1) { // We expect label as a key value pair in a single Map
      labelMap = labels[0];
    }
    try {
      // use jobs.query if possible
      if (isFastQuerySupported()) {
        logger.log(Level.INFO, "\n Using Fast Query Path");
        final String projectId = bigQueryOptions.getProjectId();
        final QueryRequest queryRequest =
            createQueryRequest(connectionSettings, sql, parameters, labelMap);
        return queryRpc(projectId, queryRequest, sql, parameters != null);
      }
      // use jobs.insert otherwise
      logger.log(Level.INFO, "\n Not Using Fast Query Path, using jobs.insert");
      com.google.api.services.bigquery.model.Job queryJob =
          createQueryJob(sql, connectionSettings, parameters, labelMap);
      JobId jobId = JobId.fromPb(queryJob.getJobReference());
      GetQueryResultsResponse firstPage = getQueryResultsFirstPage(jobId);
      return getResultSet(firstPage, jobId, sql, parameters != null);
    } catch (BigQueryException e) {
      throw new BigQuerySQLException(e.getMessage(), e, e.getErrors());
    }
  }
  /**
   * Execute a SQL statement that returns a single ResultSet and returns a ListenableFuture to
   * process the response asynchronously.
   *
   * 

Example of running a query. * *

   * {
   *   @code
   *  ConnectionSettings connectionSettings =
   *        ConnectionSettings.newBuilder()
   *            .setUseReadAPI(true)
   *            .build();
   *   Connection connection = bigquery.createConnection(connectionSettings);
   *   String selectQuery = "SELECT corpus FROM `bigquery-public-data.samples.shakespeare` GROUP BY corpus;";
   * ListenableFuture executeSelectFuture = connection.executeSelectAsync(selectQuery);
   * ExecuteSelectResponse executeSelectRes = executeSelectFuture.get();
   *
   *  if(!executeSelectRes.getIsSuccessful()){
   * throw executeSelectRes.getBigQuerySQLException();
   * }
   *
   *  BigQueryResult bigQueryResult = executeSelectRes.getBigQueryResult();
   * ResultSet rs = bigQueryResult.getResultSet();
   * while (rs.next()) {
   * System.out.println(rs.getString(1));
   * }
   *
   * 
* * @param sql a static SQL SELECT statement * @return a ListenableFuture that is used to get the data produced by the query * @throws BigQuerySQLException upon failure */ @BetaApi @Override public ListenableFuture executeSelectAsync(String sql) throws BigQuerySQLException { return getExecuteSelectFuture(sql, null); } /** This method calls the overloaded executeSelect(...) methods and returns a Future */ private ListenableFuture getExecuteSelectFuture( String sql, List parameters, Map... labels) throws BigQuerySQLException { ExecutorService execService = Executors.newFixedThreadPool( 2); // two fixed threads. One for the async operation and the other for processing the // callback ListeningExecutorService lExecService = MoreExecutors.listeningDecorator(execService); ListenableFuture executeSelectFuture = lExecService.submit( () -> { try { return ExecuteSelectResponse.newBuilder() .setResultSet( this.executeSelect( sql, parameters, labels)) // calling the overloaded executeSelect method, it takes care // of null parameters and labels .setIsSuccessful(true) .build(); } catch (BigQuerySQLException ex) { return ExecuteSelectResponse .newBuilder() // passing back the null result with isSuccessful set to false .setIsSuccessful(false) .setBigQuerySQLException(ex) .build(); } }); Futures.addCallback( executeSelectFuture, new FutureCallback() { public void onSuccess(ExecuteSelectResponse result) { execService.shutdownNow(); // shutdown the executor service as we do not need it } public void onFailure(Throwable t) { logger.log( Level.WARNING, "\n" + String.format( "Async task failed or cancelled with error %s", t.getMessage())); try { close(); // attempt to stop the execution as the developer might have called // Future.cancel() } catch (BigQuerySQLException e) { logger.log( Level.WARNING, "\n" + String.format("Exception while closing the connection %s", e.getMessage())); } execService.shutdownNow(); // shutdown the executor service as we do not need it } }, execService); return executeSelectFuture; } /** * Execute a SQL statement that returns a single ResultSet and returns a ListenableFuture to * process the response asynchronously. * *

Example of running a query. * *

   * {
   *   @code
   *  ConnectionSettings connectionSettings =
   *        ConnectionSettings.newBuilder()
   *            ..setUseReadAPI(true)
   *            .build();
   *   Connection connection = bigquery.createConnection(connectionSettings);
   *     String selectQuery =
   *         "SELECT TimestampField, StringField, BooleanField FROM "
   *             + MY_TABLE
   *             + " WHERE StringField = @stringParam"
   *             + " AND IntegerField IN UNNEST(@integerList)";
   *     QueryParameterValue stringParameter = QueryParameterValue.string("stringValue");
   *     QueryParameterValue intArrayParameter =
   *         QueryParameterValue.array(new Integer[] {3, 4}, Integer.class);
   *     Parameter stringParam =
   *         Parameter.newBuilder().setName("stringParam").setValue(stringParameter).build();
   *     Parameter intArrayParam =
   *         Parameter.newBuilder().setName("integerList").setValue(intArrayParameter).build();
   *     List parameters = ImmutableList.of(stringParam, intArrayParam);
   *
   *     ListenableFuture executeSelectFut =
   *         connection.executeSelectAsync(selectQuery, parameters);
   * ExecuteSelectResponse executeSelectRes = executeSelectFuture.get();
   *
   *  if(!executeSelectRes.getIsSuccessful()){
   * throw executeSelectRes.getBigQuerySQLException();
   * }
   *
   *  BigQueryResult bigQueryResult = executeSelectRes.getBigQueryResult();
   * ResultSet rs = bigQueryResult.getResultSet();
   * while (rs.next()) {
   * System.out.println(rs.getString(1));
   * }
   *
   * 
* * @param sql SQL SELECT query * @param parameters named or positional parameters. The set of query parameters must either be * all positional or all named parameters. * @param labels (optional) the labels associated with this query. You can use these to organize * and group your query jobs. Label keys and values can be no longer than 63 characters, can * only contain lowercase letters, numeric characters, underscores and dashes. International * characters are allowed. Label values are optional and Label is a Varargs. You should pass * all the Labels in a single Map .Label keys must start with a letter and each label in the * list must have a different key. * @return a ListenableFuture that is used to get the data produced by the query * @throws BigQuerySQLException upon failure */ @BetaApi @Override public ListenableFuture executeSelectAsync( String sql, List parameters, Map... labels) throws BigQuerySQLException { return getExecuteSelectFuture(sql, parameters, labels); } @VisibleForTesting BigQueryResult getResultSet( GetQueryResultsResponse firstPage, JobId jobId, String sql, Boolean hasQueryParameters) { if (firstPage.getTotalRows().compareTo(BigInteger.ZERO) > 0) { return getSubsequentQueryResultsWithJob( firstPage.getTotalRows().longValue(), (long) firstPage.getRows().size(), jobId, firstPage, hasQueryParameters); } return new BigQueryResultImpl(Schema.fromPb(firstPage.getSchema()), 0, null, null); } static class EndOfFieldValueList extends AbstractList< FieldValue> { // A reference of this class is used as a token to inform the thread // consuming `buffer` BigQueryResultImpl that we have run out of records @Override public FieldValue get(int index) { return null; } @Override public int size() { return 0; } } private BigQueryResult queryRpc( final String projectId, final QueryRequest queryRequest, String sql, Boolean hasQueryParameters) { com.google.api.services.bigquery.model.QueryResponse results; try { results = BigQueryRetryHelper.runWithRetries( () -> bigQueryRpc.queryRpc(projectId, queryRequest), bigQueryOptions.getRetrySettings(), BigQueryBaseService.BIGQUERY_EXCEPTION_HANDLER, bigQueryOptions.getClock(), retryConfig); } catch (BigQueryRetryHelper.BigQueryRetryHelperException e) { throw BigQueryException.translateAndThrow(e); } if (results.getErrors() != null) { List bigQueryErrors = results.getErrors().stream() .map(BigQueryError.FROM_PB_FUNCTION) .collect(Collectors.toList()); // Throwing BigQueryException since there may be no JobId, and we want to stay consistent // with the case where there is an HTTP error throw new BigQueryException(bigQueryErrors); } // Query finished running and we can paginate all the results if (results.getJobComplete() && results.getSchema() != null) { return processQueryResponseResults(results); } else { // Query is long-running (> 10s) and hasn't completed yet, or query completed but didn't // return the schema, fallback to jobs.insert path. Some operations don't return the schema // and can be optimized here, but this is left as future work. Long totalRows = results.getTotalRows() == null ? null : results.getTotalRows().longValue(); Long pageRows = results.getRows() == null ? null : (long) (results.getRows().size()); logger.log( Level.WARNING, "\n" + String.format( "results.getJobComplete(): %s, isSchemaNull: %s , totalRows: %s, pageRows: %s", results.getJobComplete(), results.getSchema() == null, totalRows, pageRows)); JobId jobId = JobId.fromPb(results.getJobReference()); GetQueryResultsResponse firstPage = getQueryResultsFirstPage(jobId); return getSubsequentQueryResultsWithJob( totalRows, pageRows, jobId, firstPage, hasQueryParameters); } } @VisibleForTesting BigQueryResultStats getBigQueryResultSetStats(JobId jobId) { // Create GetQueryResultsResponse query statistics Job queryJob = getQueryJobRpc(jobId); QueryStatistics queryStatistics = queryJob.getStatistics(); SessionInfo sessionInfo = queryStatistics.getSessionInfo() == null ? null : queryStatistics.getSessionInfo(); return new BigQueryResultStatsImpl(queryStatistics, sessionInfo); } /* This method processed the first page of GetQueryResultsResponse and then it uses tabledata.list */ @VisibleForTesting BigQueryResult tableDataList(GetQueryResultsResponse firstPage, JobId jobId) { Schema schema; long numRows; schema = Schema.fromPb(firstPage.getSchema()); numRows = firstPage.getTotalRows().longValue(); BigQueryResultStats bigQueryResultStats = getBigQueryResultSetStats(jobId); // Keeps the deserialized records at the row level, which is consumed by BigQueryResult bufferFvl = new LinkedBlockingDeque<>(getBufferSize()); // Keeps the parsed FieldValueLists BlockingQueue, Boolean>> pageCache = new LinkedBlockingDeque<>( getPageCacheSize(connectionSettings.getNumBufferedRows(), schema)); // Keeps the raw RPC responses BlockingQueue> rpcResponseQueue = new LinkedBlockingDeque<>( getPageCacheSize(connectionSettings.getNumBufferedRows(), schema)); runNextPageTaskAsync(firstPage.getPageToken(), getDestinationTable(jobId), rpcResponseQueue); parseRpcDataAsync( firstPage.getRows(), schema, pageCache, rpcResponseQueue); // parses data on a separate thread, thus maximising processing // throughput populateBufferAsync( rpcResponseQueue, pageCache, bufferFvl); // spawns a thread to populate the buffer // This will work for pagination as well, as buffer is getting updated asynchronously return new BigQueryResultImpl>( schema, numRows, bufferFvl, bigQueryResultStats); } @VisibleForTesting BigQueryResult processQueryResponseResults( com.google.api.services.bigquery.model.QueryResponse results) { Schema schema; long numRows; schema = Schema.fromPb(results.getSchema()); numRows = results.getTotalRows() == null ? 0 : results.getTotalRows().longValue(); // in case of DML or DDL // QueryResponse only provides cache hits, dmlStats, and sessionInfo as query processing // statistics DmlStats dmlStats = results.getDmlStats() == null ? null : DmlStats.fromPb(results.getDmlStats()); Boolean cacheHit = results.getCacheHit(); QueryStatistics queryStatistics = QueryStatistics.newBuilder().setDmlStats(dmlStats).setCacheHit(cacheHit).build(); // We cannot directly set sessionInfo in QueryStatistics SessionInfo sessionInfo = results.getSessionInfo() == null ? null : JobStatistics.SessionInfo.fromPb(results.getSessionInfo()); BigQueryResultStats bigQueryResultStats = new BigQueryResultStatsImpl(queryStatistics, sessionInfo); bufferFvl = new LinkedBlockingDeque<>(getBufferSize()); BlockingQueue, Boolean>> pageCache = new LinkedBlockingDeque<>( getPageCacheSize(connectionSettings.getNumBufferedRows(), schema)); BlockingQueue> rpcResponseQueue = new LinkedBlockingDeque<>( getPageCacheSize(connectionSettings.getNumBufferedRows(), schema)); JobId jobId = JobId.fromPb(results.getJobReference()); // Thread to make rpc calls to fetch data from the server runNextPageTaskAsync(results.getPageToken(), getDestinationTable(jobId), rpcResponseQueue); // Thread to parse data received from the server to client library objects parseRpcDataAsync(results.getRows(), schema, pageCache, rpcResponseQueue); // Thread to populate the buffer (a blocking queue) shared with the consumer populateBufferAsync(rpcResponseQueue, pageCache, bufferFvl); return new BigQueryResultImpl>( schema, numRows, bufferFvl, bigQueryResultStats); } @VisibleForTesting void runNextPageTaskAsync( String firstPageToken, TableId destinationTable, BlockingQueue> rpcResponseQueue) { // This thread makes the RPC calls and paginates Runnable nextPageTask = () -> { String pageToken = firstPageToken; // results.getPageToken(); try { while (pageToken != null) { // paginate for non null token if (Thread.currentThread().isInterrupted() || queryTaskExecutor.isShutdown()) { // do not process further pages and shutdown logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Interrupted @ runNextPageTaskAsync"); break; } TableDataList tabledataList = tableDataListRpc(destinationTable, pageToken); pageToken = tabledataList.getPageToken(); rpcResponseQueue.put( Tuple.of( tabledataList, true)); // this will be parsed asynchronously without blocking the current // thread } rpcResponseQueue.put( Tuple.of( null, false)); // this will stop the parseDataTask as well when the pagination // completes } catch (Exception e) { throw new BigQueryException(0, e.getMessage(), e); } // We cannot do queryTaskExecutor.shutdownNow() here as populate buffer method may not // have finished processing the records and even that will be interrupted }; queryTaskExecutor.execute(nextPageTask); } /* This method takes TableDataList from rpcResponseQueue and populates pageCache with FieldValueList */ @VisibleForTesting void parseRpcDataAsync( // com.google.api.services.bigquery.model.QueryResponse results, List tableRows, Schema schema, BlockingQueue, Boolean>> pageCache, BlockingQueue> rpcResponseQueue) { // parse and put the first page in the pageCache before the other pages are parsed from the RPC // calls Iterable firstFieldValueLists = getIterableFieldValueList(tableRows, schema); try { pageCache.put( Tuple.of(firstFieldValueLists, true)); // this is the first page which we have received. } catch (InterruptedException e) { logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Interrupted @ parseRpcDataAsync"); } // rpcResponseQueue will get null tuple if Cancel method is called, so no need to explicitly use // thread interrupt here Runnable parseDataTask = () -> { try { boolean hasMorePages = true; while (hasMorePages) { if (Thread.currentThread().isInterrupted() || queryTaskExecutor.isShutdown()) { // do not process further data and shutdown logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Interrupted @ parseRpcDataAsync"); break; } // no interrupt received till this point, continue processing Tuple rpcResponse = rpcResponseQueue.take(); TableDataList tabledataList = rpcResponse.x(); hasMorePages = rpcResponse.y(); if (tabledataList != null) { Iterable fieldValueLists = getIterableFieldValueList(tabledataList.getRows(), schema); // Parse pageCache.put(Tuple.of(fieldValueLists, true)); } } } catch (InterruptedException e) { logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Interrupted @ parseRpcDataAsync", e); // Thread might get interrupted while calling the Cancel method, which is // expected, so logging this instead of throwing the exception back } try { pageCache.put(Tuple.of(null, false)); // no further pages, graceful exit scenario } catch (InterruptedException e) { logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Interrupted @ parseRpcDataAsync", e); // Thread might get interrupted while calling the Cancel method, which is // expected, so logging this instead of throwing the exception back } // We cannot do queryTaskExecutor.shutdownNow() here as populate buffer method may not // have finished processing the records and even that will be interrupted }; queryTaskExecutor.execute(parseDataTask); } @VisibleForTesting void populateBufferAsync( BlockingQueue> rpcResponseQueue, BlockingQueue, Boolean>> pageCache, BlockingQueue> buffer) { Runnable populateBufferRunnable = () -> { // producer thread populating the buffer Iterable fieldValueLists = null; boolean hasRows = true; // as we have to process the first page while (hasRows) { try { Tuple, Boolean> nextPageTuple = pageCache.take(); hasRows = nextPageTuple.y(); fieldValueLists = nextPageTuple.x(); } catch (InterruptedException e) { logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Interrupted", e); // Thread might get interrupted while calling the Cancel method, which is // expected, so logging this instead of throwing the exception back break; } if (Thread.currentThread().isInterrupted() || queryTaskExecutor.isShutdown() || fieldValueLists == null) { // do not process further pages and shutdown (outerloop) break; } for (FieldValueList fieldValueList : fieldValueLists) { try { if (Thread.currentThread().isInterrupted() || queryTaskExecutor .isShutdown()) { // do not process further pages and shutdown (inner loop) break; } buffer.put(fieldValueList); } catch (InterruptedException e) { throw new BigQueryException(0, e.getMessage(), e); } } } try { buffer.put( new EndOfFieldValueList()); // All the pages has been processed, put this marker } catch (InterruptedException e) { logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Interrupted @ populateBufferAsync", e); } finally { queryTaskExecutor .shutdownNow(); // Shutdown the thread pool. All the records are now processed } }; queryTaskExecutor.execute(populateBufferRunnable); } /** * In an interrupt scenario, like when the background threads are still working and the user calls * `connection.close() then we need to add an End of Stream flag in the buffer so that the * `ResultSet.next()` stops advancing the cursor. We cannot rely on the `populateBufferAsync` * method to do this as the `BlockingQueue.put()` call will error out after the interrupt is * triggerred */ @InternalApi void flagEndOfStream() { // package-private try { if (bufferFvl != null) { // that is tabledata.list endpoint is used bufferFvl.put( new EndOfFieldValueList()); // All the pages has been processed, put this marker } else if (bufferRow != null) { bufferRow.put( new BigQueryResultImpl.Row( null, true)); // All the pages has been processed, put this marker } else { logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Could not flag End of Stream, both the buffer types are null. This might happen when the connection is close without executing a query"); } } catch (InterruptedException e) { logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Interrupted @ flagEndOfStream", e); } } /* Helper method that parse and populate a page with TableRows */ private static Iterable getIterableFieldValueList( Iterable tableDataPb, final Schema schema) { return ImmutableList.copyOf( Iterables.transform( tableDataPb != null ? tableDataPb : ImmutableList.of(), new Function() { final FieldList fields = schema != null ? schema.getFields() : null; @Override public FieldValueList apply(TableRow rowPb) { return FieldValueList.fromPb(rowPb.getF(), fields); } })); } /* Helper method that determines the optimal number of caches pages to improve read performance */ @VisibleForTesting int getPageCacheSize(Integer numBufferedRows, Schema schema) { final int MIN_CACHE_SIZE = 3; // Min number of pages to cache final int MAX_CACHE_SIZE = 20; // //Min number of pages to cache int numColumns = schema.getFields().size(); int numCachedPages; long numCachedRows = numBufferedRows == null ? 0 : numBufferedRows.longValue(); // TODO: Further enhance this logic depending on customer feedback on memory consumption if (numCachedRows > 10000) { numCachedPages = 2; // the size of numBufferedRows is quite large and as per our tests we should be able to // do enough even with low } else if (numColumns > 15 && numCachedRows > 5000) { // too many fields are being read, setting the page size on the lower end numCachedPages = 3; } else if (numCachedRows < 2000 && numColumns < 15) { // low pagesize with fewer number of columns, we can cache more pages numCachedPages = 20; } else { // default - under 10K numCachedRows with any number of columns numCachedPages = 5; } return numCachedPages < MIN_CACHE_SIZE ? MIN_CACHE_SIZE : (Math.min( numCachedPages, MAX_CACHE_SIZE)); // numCachedPages should be between the defined min and max } /* Returns query results using either tabledata.list or the high throughput Read API */ @VisibleForTesting BigQueryResult getSubsequentQueryResultsWithJob( Long totalRows, Long pageRows, JobId jobId, GetQueryResultsResponse firstPage, Boolean hasQueryParameters) { TableId destinationTable = getDestinationTable(jobId); return useReadAPI(totalRows, pageRows, Schema.fromPb(firstPage.getSchema()), hasQueryParameters) ? highThroughPutRead( destinationTable, firstPage.getTotalRows().longValue(), Schema.fromPb(firstPage.getSchema()), getBigQueryResultSetStats( jobId)) // discord first page and stream the entire BigQueryResult using // the Read API : tableDataList(firstPage, jobId); } /* Returns query results using either tabledata.list or the high throughput Read API */ @VisibleForTesting BigQueryResult getSubsequentQueryResultsWithJob( Long totalRows, Long pageRows, JobId jobId, GetQueryResultsResponse firstPage, Schema schema, Boolean hasQueryParameters) { TableId destinationTable = getDestinationTable(jobId); return useReadAPI(totalRows, pageRows, schema, hasQueryParameters) ? highThroughPutRead( destinationTable, totalRows == null ? -1L : totalRows, // totalRows is null when the job is still running. TODO: Check if // any workaround is possible schema, getBigQueryResultSetStats( jobId)) // discord first page and stream the entire BigQueryResult using // the Read API : tableDataList(firstPage, jobId); } /* Returns Job from jobId by calling the jobs.get API */ private Job getQueryJobRpc(JobId jobId) { final JobId completeJobId = jobId .setProjectId(bigQueryOptions.getProjectId()) .setLocation( jobId.getLocation() == null && bigQueryOptions.getLocation() != null ? bigQueryOptions.getLocation() : jobId.getLocation()); com.google.api.services.bigquery.model.Job jobPb; try { jobPb = runWithRetries( () -> bigQueryRpc.getQueryJob( completeJobId.getProject(), completeJobId.getJob(), completeJobId.getLocation()), bigQueryOptions.getRetrySettings(), BigQueryBaseService.BIGQUERY_EXCEPTION_HANDLER, bigQueryOptions.getClock()); if (bigQueryOptions.getThrowNotFound() && jobPb == null) { throw new BigQueryException(HTTP_NOT_FOUND, "Query job not found"); } } catch (RetryHelper.RetryHelperException e) { throw BigQueryException.translateAndThrow(e); } return Job.fromPb(bigQueryOptions.getService(), jobPb); } /* Returns the destinationTable from jobId by calling jobs.get API */ @VisibleForTesting TableId getDestinationTable(JobId jobId) { Job job = getQueryJobRpc(jobId); return ((QueryJobConfiguration) job.getConfiguration()).getDestinationTable(); } @VisibleForTesting TableDataList tableDataListRpc(TableId destinationTable, String pageToken) { try { final TableId completeTableId = destinationTable.setProjectId( Strings.isNullOrEmpty(destinationTable.getProject()) ? bigQueryOptions.getProjectId() : destinationTable.getProject()); TableDataList results = runWithRetries( () -> bigQueryOptions .getBigQueryRpcV2() .listTableDataWithRowLimit( completeTableId.getProject(), completeTableId.getDataset(), completeTableId.getTable(), connectionSettings.getMaxResultPerPage(), pageToken), bigQueryOptions.getRetrySettings(), BigQueryBaseService.BIGQUERY_EXCEPTION_HANDLER, bigQueryOptions.getClock()); return results; } catch (RetryHelper.RetryHelperException e) { throw BigQueryException.translateAndThrow(e); } } @VisibleForTesting BigQueryResult highThroughPutRead( TableId destinationTable, long totalRows, Schema schema, BigQueryResultStats stats) { try { if (bqReadClient == null) { // if the read client isn't already initialized. Not thread safe. BigQueryReadSettings settings = BigQueryReadSettings.newBuilder() .setCredentialsProvider( FixedCredentialsProvider.create(bigQueryOptions.getCredentials())) .build(); bqReadClient = BigQueryReadClient.create(settings); } String parent = String.format("projects/%s", destinationTable.getProject()); String srcTable = String.format( "projects/%s/datasets/%s/tables/%s", destinationTable.getProject(), destinationTable.getDataset(), destinationTable.getTable()); // Read all the columns if the source table (temp table) and stream the data back in Arrow // format ReadSession.Builder sessionBuilder = ReadSession.newBuilder().setTable(srcTable).setDataFormat(DataFormat.ARROW); CreateReadSessionRequest.Builder builder = CreateReadSessionRequest.newBuilder() .setParent(parent) .setReadSession(sessionBuilder) .setMaxStreamCount(1) // Currently just one stream is allowed // DO a regex check using order by and use multiple streams ; ReadSession readSession = bqReadClient.createReadSession(builder.build()); bufferRow = new LinkedBlockingDeque<>(getBufferSize()); Map arrowNameToIndex = new HashMap<>(); // deserialize and populate the buffer async, so that the client isn't blocked processArrowStreamAsync( readSession, bufferRow, new ArrowRowReader(readSession.getArrowSchema(), arrowNameToIndex), schema); logger.log(Level.INFO, "\n Using BigQuery Read API"); return new BigQueryResultImpl(schema, totalRows, bufferRow, stats); } catch (IOException e) { throw BigQueryException.translateAndThrow(e); } } private void processArrowStreamAsync( ReadSession readSession, BlockingQueue buffer, ArrowRowReader reader, Schema schema) { Runnable arrowStreamProcessor = () -> { try { // Use the first stream to perform reading. String streamName = readSession.getStreams(0).getName(); ReadRowsRequest readRowsRequest = ReadRowsRequest.newBuilder().setReadStream(streamName).build(); // Process each block of rows as they arrive and decode using our simple row reader. com.google.api.gax.rpc.ServerStream stream = bqReadClient.readRowsCallable().call(readRowsRequest); for (ReadRowsResponse response : stream) { if (Thread.currentThread().isInterrupted() || queryTaskExecutor.isShutdown()) { // do not process and shutdown break; } reader.processRows(response.getArrowRecordBatch(), buffer, schema); } } catch (Exception e) { throw BigQueryException.translateAndThrow(e); } finally { // logic needed for graceful shutdown // marking end of stream try { buffer.put( new BigQueryResultImpl.Row( null, true)); // All the pages has been processed, put this marker } catch (InterruptedException e) { logger.log( Level.WARNING, "\n" + Thread.currentThread().getName() + " Interrupted @ markLast", e); } queryTaskExecutor.shutdownNow(); // Shutdown the thread pool } }; queryTaskExecutor.execute(arrowStreamProcessor); } private class ArrowRowReader implements AutoCloseable { // TODO: Update to recent version of Arrow to avoid memoryleak BufferAllocator allocator = new RootAllocator(Long.MAX_VALUE); // Decoder object will be reused to avoid re-allocation and too much garbage collection. private final VectorSchemaRoot root; private final VectorLoader loader; private ArrowRowReader(ArrowSchema arrowSchema, Map arrowNameToIndex) throws IOException { org.apache.arrow.vector.types.pojo.Schema schema = MessageSerializer.deserializeSchema( new org.apache.arrow.vector.ipc.ReadChannel( new ByteArrayReadableSeekableByteChannel( arrowSchema.getSerializedSchema().toByteArray()))); List vectors = new ArrayList<>(); List fields = schema.getFields(); for (int i = 0; i < fields.size(); i++) { vectors.add(fields.get(i).createVector(allocator)); arrowNameToIndex.put( fields.get(i).getName(), i); // mapping for getting against the field name in the result set } root = new VectorSchemaRoot(vectors); loader = new VectorLoader(root); } /** @param batch object returned from the ReadRowsResponse. */ private void processRows( ArrowRecordBatch batch, BlockingQueue buffer, Schema schema) throws IOException { // deserialize the values and consume the hash of the values try { org.apache.arrow.vector.ipc.message.ArrowRecordBatch deserializedBatch = MessageSerializer.deserializeRecordBatch( new ReadChannel( new ByteArrayReadableSeekableByteChannel( batch.getSerializedRecordBatch().toByteArray())), allocator); loader.load(deserializedBatch); // Release buffers from batch (they are still held in the vectors in root). deserializedBatch.close(); // Parse the vectors using BQ Schema. Deserialize the data at the row level and add it to // the // buffer FieldList fields = schema.getFields(); for (int rowNum = 0; rowNum < root.getRowCount(); rowNum++) { // for the given number of rows in the batch if (Thread.currentThread().isInterrupted() || queryTaskExecutor.isShutdown()) { // do not process and shutdown break; // exit the loop, root will be cleared in the finally block } Map curRow = new HashMap<>(); for (int col = 0; col < fields.size(); col++) { // iterate all the vectors for a given row com.google.cloud.bigquery.Field field = fields.get(col); FieldVector curFieldVec = root.getVector( field.getName()); // can be accessed using the index or Vector/column name curRow.put(field.getName(), curFieldVec.getObject(rowNum)); // Added the raw value } buffer.put(new BigQueryResultImpl.Row(curRow)); } root.clear(); } catch (RuntimeException | InterruptedException e) { throw BigQueryException.translateAndThrow(e); } finally { try { root.clear(); } catch (RuntimeException e) { logger.log(Level.WARNING, "\n Error while clearing VectorSchemaRoot ", e); } } } @Override public void close() { root.close(); allocator.close(); } } /*Returns just the first page of GetQueryResultsResponse using the jobId*/ @VisibleForTesting GetQueryResultsResponse getQueryResultsFirstPage(JobId jobId) { JobId completeJobId = jobId .setProjectId(bigQueryOptions.getProjectId()) .setLocation( jobId.getLocation() == null && bigQueryOptions.getLocation() != null ? bigQueryOptions.getLocation() : jobId.getLocation()); // Implementing logic to poll the Job's status using getQueryResults as // we do not get rows, rows count and schema unless the job is complete // Ref: b/241134681 // This logic relies on backend for poll and wait.BigQuery guarantees that jobs make forward // progress (a job won't get stuck in pending forever). boolean jobComplete = false; GetQueryResultsResponse results = null; long timeoutMs = 10000; // defaulting to 10seconds. while (!jobComplete) { try { results = BigQueryRetryHelper.runWithRetries( () -> bigQueryRpc.getQueryResultsWithRowLimit( completeJobId.getProject(), completeJobId.getJob(), completeJobId.getLocation(), connectionSettings.getMaxResultPerPage(), timeoutMs), bigQueryOptions.getRetrySettings(), BigQueryBaseService.BIGQUERY_EXCEPTION_HANDLER, bigQueryOptions.getClock(), retryConfig); if (results.getErrors() != null) { List bigQueryErrors = results.getErrors().stream() .map(BigQueryError.FROM_PB_FUNCTION) .collect(Collectors.toList()); // Throwing BigQueryException since there may be no JobId, and we want to stay consistent // with the case where there is a HTTP error throw new BigQueryException(bigQueryErrors); } } catch (BigQueryRetryHelper.BigQueryRetryHelperException e) { logger.log(Level.WARNING, "\n Error occurred while calling getQueryResultsWithRowLimit", e); throw BigQueryException.translateAndThrow(e); } jobComplete = results.getJobComplete(); // This log msg at Level.FINE might indicate that the job is still running and not stuck for // very long running jobs. logger.log( Level.FINE, String.format( "jobComplete: %s , Polling getQueryResults with timeoutMs: %s", jobComplete, timeoutMs)); } return results; } @VisibleForTesting boolean isFastQuerySupported() { // TODO: add regex logic to check for scripting return connectionSettings.getClustering() == null && connectionSettings.getCreateDisposition() == null && connectionSettings.getDestinationEncryptionConfiguration() == null && connectionSettings.getDestinationTable() == null && connectionSettings.getJobTimeoutMs() == null && connectionSettings.getMaximumBillingTier() == null && connectionSettings.getPriority() == null && connectionSettings.getRangePartitioning() == null && connectionSettings.getSchemaUpdateOptions() == null && connectionSettings.getTableDefinitions() == null && connectionSettings.getTimePartitioning() == null && connectionSettings.getUserDefinedFunctions() == null && connectionSettings.getWriteDisposition() == null; } @VisibleForTesting boolean useReadAPI(Long totalRows, Long pageRows, Schema schema, Boolean hasQueryParameters) { // Read API does not yet support Interval Type or QueryParameters if (containsIntervalType(schema) || hasQueryParameters) { logger.log(Level.INFO, "\n Schema has IntervalType, or QueryParameters. Disabling ReadAPI"); return false; } if (totalRows == null || pageRows == null) { return connectionSettings.getUseReadAPI(); } if (Boolean.TRUE.equals(connectionSettings.getUseReadAPI())) { long resultRatio = totalRows / pageRows; return resultRatio >= connectionSettings.getTotalToPageRowCountRatio() && totalRows > connectionSettings.getMinResultSize(); } else { return false; } } // Does a BFS iteration to find out if there's an interval type in the schema. Implementation to // be used until ReadAPI supports IntervalType private boolean containsIntervalType(Schema schema) { Queue fields = new LinkedList(schema.getFields()); while (!fields.isEmpty()) { com.google.cloud.bigquery.Field curField = fields.poll(); if (curField.getType().getStandardType() == StandardSQLTypeName.INTERVAL) { return true; } else if (curField.getType().getStandardType() == StandardSQLTypeName.STRUCT || curField.getType().getStandardType() == StandardSQLTypeName.ARRAY) { fields.addAll(curField.getSubFields()); } } return false; } // Used for job.query API endpoint @VisibleForTesting QueryRequest createQueryRequest( ConnectionSettings connectionSettings, String sql, List queryParameters, Map labels) { QueryRequest content = new QueryRequest(); String requestId = UUID.randomUUID().toString(); if (connectionSettings.getConnectionProperties() != null) { content.setConnectionProperties( connectionSettings.getConnectionProperties().stream() .map(ConnectionProperty.TO_PB_FUNCTION) .collect(Collectors.toList())); } if (connectionSettings.getDefaultDataset() != null) { content.setDefaultDataset(connectionSettings.getDefaultDataset().toPb()); } if (connectionSettings.getMaximumBytesBilled() != null) { content.setMaximumBytesBilled(connectionSettings.getMaximumBytesBilled()); } if (connectionSettings.getMaxResults() != null) { content.setMaxResults(connectionSettings.getMaxResults()); } if (queryParameters != null) { // content.setQueryParameters(queryParameters); if (queryParameters.get(0).getName() == null) { // If query parameter name is unset, then assume mode is positional content.setParameterMode("POSITIONAL"); // pass query parameters List queryParametersPb = Lists.transform(queryParameters, POSITIONAL_PARAMETER_TO_PB_FUNCTION); content.setQueryParameters(queryParametersPb); } else { content.setParameterMode("NAMED"); // pass query parameters List queryParametersPb = Lists.transform(queryParameters, NAMED_PARAMETER_TO_PB_FUNCTION); content.setQueryParameters(queryParametersPb); } } if (connectionSettings.getCreateSession() != null) { content.setCreateSession(connectionSettings.getCreateSession()); } if (labels != null) { content.setLabels(labels); } content.setQuery(sql); content.setRequestId(requestId); // The new Connection interface only supports StandardSQL dialect content.setUseLegacySql(false); return content; } // Used by jobs.getQueryResults API endpoint @VisibleForTesting com.google.api.services.bigquery.model.Job createQueryJob( String sql, ConnectionSettings connectionSettings, List queryParameters, Map labels) { com.google.api.services.bigquery.model.JobConfiguration configurationPb = new com.google.api.services.bigquery.model.JobConfiguration(); JobConfigurationQuery queryConfigurationPb = new JobConfigurationQuery(); queryConfigurationPb.setQuery(sql); if (queryParameters != null) { if (queryParameters.get(0).getName() == null) { // If query parameter name is unset, then assume mode is positional queryConfigurationPb.setParameterMode("POSITIONAL"); // pass query parameters List queryParametersPb = Lists.transform(queryParameters, POSITIONAL_PARAMETER_TO_PB_FUNCTION); queryConfigurationPb.setQueryParameters(queryParametersPb); } else { queryConfigurationPb.setParameterMode("NAMED"); // pass query parameters List queryParametersPb = Lists.transform(queryParameters, NAMED_PARAMETER_TO_PB_FUNCTION); queryConfigurationPb.setQueryParameters(queryParametersPb); } } if (connectionSettings.getDestinationTable() != null) { queryConfigurationPb.setDestinationTable(connectionSettings.getDestinationTable().toPb()); } if (connectionSettings.getTableDefinitions() != null) { queryConfigurationPb.setTableDefinitions( Maps.transformValues( connectionSettings.getTableDefinitions(), ExternalTableDefinition.TO_EXTERNAL_DATA_FUNCTION)); } if (connectionSettings.getUserDefinedFunctions() != null) { queryConfigurationPb.setUserDefinedFunctionResources( connectionSettings.getUserDefinedFunctions().stream() .map(UserDefinedFunction.TO_PB_FUNCTION) .collect(Collectors.toList())); } if (connectionSettings.getCreateDisposition() != null) { queryConfigurationPb.setCreateDisposition( connectionSettings.getCreateDisposition().toString()); } if (connectionSettings.getWriteDisposition() != null) { queryConfigurationPb.setWriteDisposition(connectionSettings.getWriteDisposition().toString()); } if (connectionSettings.getDefaultDataset() != null) { queryConfigurationPb.setDefaultDataset(connectionSettings.getDefaultDataset().toPb()); } if (connectionSettings.getPriority() != null) { queryConfigurationPb.setPriority(connectionSettings.getPriority().toString()); } if (connectionSettings.getAllowLargeResults() != null) { queryConfigurationPb.setAllowLargeResults(connectionSettings.getAllowLargeResults()); } if (connectionSettings.getUseQueryCache() != null) { queryConfigurationPb.setUseQueryCache(connectionSettings.getUseQueryCache()); } if (connectionSettings.getFlattenResults() != null) { queryConfigurationPb.setFlattenResults(connectionSettings.getFlattenResults()); } if (connectionSettings.getMaximumBillingTier() != null) { queryConfigurationPb.setMaximumBillingTier(connectionSettings.getMaximumBillingTier()); } if (connectionSettings.getMaximumBytesBilled() != null) { queryConfigurationPb.setMaximumBytesBilled(connectionSettings.getMaximumBytesBilled()); } if (connectionSettings.getSchemaUpdateOptions() != null) { ImmutableList.Builder schemaUpdateOptionsBuilder = new ImmutableList.Builder<>(); for (JobInfo.SchemaUpdateOption schemaUpdateOption : connectionSettings.getSchemaUpdateOptions()) { schemaUpdateOptionsBuilder.add(schemaUpdateOption.name()); } queryConfigurationPb.setSchemaUpdateOptions(schemaUpdateOptionsBuilder.build()); } if (connectionSettings.getDestinationEncryptionConfiguration() != null) { queryConfigurationPb.setDestinationEncryptionConfiguration( connectionSettings.getDestinationEncryptionConfiguration().toPb()); } if (connectionSettings.getTimePartitioning() != null) { queryConfigurationPb.setTimePartitioning(connectionSettings.getTimePartitioning().toPb()); } if (connectionSettings.getClustering() != null) { queryConfigurationPb.setClustering(connectionSettings.getClustering().toPb()); } if (connectionSettings.getRangePartitioning() != null) { queryConfigurationPb.setRangePartitioning(connectionSettings.getRangePartitioning().toPb()); } if (connectionSettings.getConnectionProperties() != null) { queryConfigurationPb.setConnectionProperties( connectionSettings.getConnectionProperties().stream() .map(ConnectionProperty.TO_PB_FUNCTION) .collect(Collectors.toList())); } if (connectionSettings.getCreateSession() != null) { queryConfigurationPb.setCreateSession(connectionSettings.getCreateSession()); } if (connectionSettings.getJobTimeoutMs() != null) { configurationPb.setJobTimeoutMs(connectionSettings.getJobTimeoutMs()); } if (labels != null) { configurationPb.setLabels(labels); } // The new Connection interface only supports StandardSQL dialect queryConfigurationPb.setUseLegacySql(false); configurationPb.setQuery(queryConfigurationPb); com.google.api.services.bigquery.model.Job jobPb = JobInfo.of(QueryJobConfiguration.fromPb(configurationPb)).toPb(); com.google.api.services.bigquery.model.Job queryJob; try { queryJob = BigQueryRetryHelper.runWithRetries( () -> bigQueryRpc.createJobForQuery(jobPb), bigQueryOptions.getRetrySettings(), BigQueryBaseService.BIGQUERY_EXCEPTION_HANDLER, bigQueryOptions.getClock(), retryConfig); } catch (BigQueryRetryHelper.BigQueryRetryHelperException e) { logger.log(Level.WARNING, "\n Error occurred while calling createJobForQuery", e); throw BigQueryException.translateAndThrow(e); } logger.log(Level.INFO, "\n Query job created"); return queryJob; } // Used by dryRun @VisibleForTesting com.google.api.services.bigquery.model.Job createDryRunJob(String sql) { com.google.api.services.bigquery.model.JobConfiguration configurationPb = new com.google.api.services.bigquery.model.JobConfiguration(); configurationPb.setDryRun(true); JobConfigurationQuery queryConfigurationPb = new JobConfigurationQuery(); String parameterMode = sql.contains("?") ? "POSITIONAL" : "NAMED"; queryConfigurationPb.setParameterMode(parameterMode); queryConfigurationPb.setQuery(sql); // UndeclaredQueryParameter is only supported in StandardSQL queryConfigurationPb.setUseLegacySql(false); if (connectionSettings.getDefaultDataset() != null) { queryConfigurationPb.setDefaultDataset(connectionSettings.getDefaultDataset().toPb()); } if (connectionSettings.getCreateSession() != null) { queryConfigurationPb.setCreateSession(connectionSettings.getCreateSession()); } configurationPb.setQuery(queryConfigurationPb); com.google.api.services.bigquery.model.Job jobPb = JobInfo.of(QueryJobConfiguration.fromPb(configurationPb)).toPb(); com.google.api.services.bigquery.model.Job dryRunJob; try { dryRunJob = BigQueryRetryHelper.runWithRetries( () -> bigQueryRpc.createJobForQuery(jobPb), bigQueryOptions.getRetrySettings(), BigQueryBaseService.BIGQUERY_EXCEPTION_HANDLER, bigQueryOptions.getClock(), retryConfig); } catch (BigQueryRetryHelper.BigQueryRetryHelperException e) { throw BigQueryException.translateAndThrow(e); } return dryRunJob; } // Convert from Parameter wrapper class to positional QueryParameter generated class private static final Function POSITIONAL_PARAMETER_TO_PB_FUNCTION = value -> { QueryParameter queryParameterPb = new QueryParameter(); queryParameterPb.setParameterValue(value.getValue().toValuePb()); queryParameterPb.setParameterType(value.getValue().toTypePb()); return queryParameterPb; }; // Convert from Parameter wrapper class to name QueryParameter generated class private static final Function NAMED_PARAMETER_TO_PB_FUNCTION = value -> { QueryParameter queryParameterPb = new QueryParameter(); queryParameterPb.setName(value.getName()); queryParameterPb.setParameterValue(value.getValue().toValuePb()); queryParameterPb.setParameterType(value.getValue().toTypePb()); return queryParameterPb; }; // Convert from QueryParameter class to the Parameter wrapper class private static final Function QUERY_PARAMETER_FROM_PB_FUNCTION = pb -> Parameter.newBuilder() .setName(pb.getName() == null ? "" : pb.getName()) .setValue(QueryParameterValue.fromPb(pb.getParameterValue(), pb.getParameterType())) .build(); }




© 2015 - 2024 Weber Informatics LLC | Privacy Policy