kylemeow commented on a change in pull request #12803: URL: https://github.com/apache/flink/pull/12803#discussion_r580043425
########## File path: flink-connectors/flink-connector-jdbc/src/main/java/org/apache/flink/connector/jdbc/internal/executor/BufferReduceStatementExecutor.java ########## @@ -0,0 +1,108 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you 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 org.apache.flink.connector.jdbc.internal.executor; + +import org.apache.flink.api.java.tuple.Tuple2; +import org.apache.flink.table.data.RowData; +import org.apache.flink.types.RowKind; + +import java.sql.Connection; +import java.sql.SQLException; +import java.util.HashMap; +import java.util.Map; +import java.util.function.Function; + +/** + * Currently, this statement executor is only used for table/sql to buffer insert/update/delete + * events, and reduce them in buffer before submit to external database. + */ +public class BufferReduceStatementExecutor implements JdbcBatchStatementExecutor<RowData> { + + private final JdbcBatchStatementExecutor<RowData> upsertExecutor; + private final JdbcBatchStatementExecutor<RowData> deleteExecutor; + private final Function<RowData, RowData> keyExtractor; + private final Function<RowData, RowData> valueTransform; + // the mapping is [KEY, <+/-, VALUE>] + private final Map<RowData, Tuple2<Boolean, RowData>> reduceBuffer = new HashMap<>(); + + public BufferReduceStatementExecutor( + JdbcBatchStatementExecutor<RowData> upsertExecutor, + JdbcBatchStatementExecutor<RowData> deleteExecutor, + Function<RowData, RowData> keyExtractor, + Function<RowData, RowData> valueTransform) { + this.upsertExecutor = upsertExecutor; + this.deleteExecutor = deleteExecutor; + this.keyExtractor = keyExtractor; + this.valueTransform = valueTransform; + } + + @Override + public void prepareStatements(Connection connection) throws SQLException { + upsertExecutor.prepareStatements(connection); + deleteExecutor.prepareStatements(connection); + } + + @Override + public void addToBatch(RowData record) throws SQLException { + RowData key = keyExtractor.apply(record); + boolean flag = changeFlag(record.getRowKind()); + RowData value = valueTransform.apply(record); // copy or not + reduceBuffer.put(key, Tuple2.of(flag, value)); + } + + /** + * Returns true if the row kind is INSERT or UPDATE_AFTER, + * returns false if the row kind is DELETE or UPDATE_BEFORE. + */ + private boolean changeFlag(RowKind rowKind) { + switch (rowKind) { + case INSERT: + case UPDATE_AFTER: + return true; + case DELETE: + case UPDATE_BEFORE: + return false; + default: + throw new UnsupportedOperationException( + String.format("Unknown row kind, the supported row kinds is: INSERT, UPDATE_BEFORE, UPDATE_AFTER," + + " DELETE, but get: %s.", rowKind)); + } + } + + @Override + public void executeBatch() throws SQLException { + // TODO: reuse the extracted key and avoid value copy in the future + for (Tuple2<Boolean, RowData> tuple : reduceBuffer.values()) { + if (tuple.f0) { + upsertExecutor.addToBatch(tuple.f1); + } else { + deleteExecutor.addToBatch(tuple.f1); + } + } + upsertExecutor.executeBatch(); + deleteExecutor.executeBatch(); Review comment: Thanks @wuchong for the timely comments. However, I have observed that some records in `deleteExecutor` contains the same primary key as the records in `upsertExecutor`, and so they are eventually deleted from the database and never appear thereafter. This problem only happens when "connector.write.flush.max-rows" is larger than 1, so I wonder if this could be a bug related to the bulk flush logic? Thanks again for your attention : ) ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org
