Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Feature-3071][CDCSOURCE] CDCSOURCE supports ensuring data order in multiple degrees of parallelism #3072

Merged
merged 1 commit into from
Jan 26, 2024
Merged

[Feature-3071][CDCSOURCE] CDCSOURCE supports ensuring data order in multiple degrees of parallelism #3072

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions dinky-cdc/dinky-cdc-core/src/main/java/org/dinky/cdc/AbstractSinkBuilder.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -146,8 +146,8 @@ protected SingleOutputStreamOperator<Map> shunt(
}

protected DataStream<Map> shunt(SingleOutputStreamOperator<Map> processOperator, Table table, OutputTag<Map> tag) {

return processOperator.getSideOutput(tag);
processOperator.forward();
return processOperator.getSideOutput(tag).forward();
}

@SuppressWarnings("rawtypes")
Expand Down
32 changes: 31 additions & 1 deletion dinky-cdc/dinky-cdc-core/src/main/java/org/dinky/cdc/sql/AbstractSqlSinkBuilder.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,7 @@
import org.dinky.utils.LogUtil;

import org.apache.flink.api.common.functions.FlatMapFunction;
import org.apache.flink.api.common.functions.Partitioner;
import org.apache.flink.api.common.typeinfo.TypeInformation;
import org.apache.flink.api.dag.Transformation;
import org.apache.flink.api.java.typeutils.RowTypeInfo;
Expand All @@ -55,6 +56,7 @@
import java.util.List;
import java.util.Map;
import java.util.Optional;
import java.util.stream.Collectors;

public abstract class AbstractSqlSinkBuilder extends AbstractSinkBuilder implements Serializable {
protected ZoneId sinkTimeZone = ZoneId.of("UTC");
Expand Down Expand Up @@ -147,7 +149,7 @@ protected void addTableSinkForTags(
buildColumn(columnNameList, columnTypeList, table.getColumns());
DataStream<Row> rowDataDataStream = buildRow(
filterOperator, columnNameList, columnTypeList, schemaTableName)
.rebalance();
.forward();
logger.info("Build {} flatMap successful...", schemaTableName);
logger.info("Start build {} sink...", schemaTableName);

Expand All @@ -166,6 +168,7 @@ protected SingleOutputStreamOperator<Map> createMapSingleOutputStreamOperator(
SingleOutputStreamOperator<Map> mapOperator =
dataStreamSource.map(x -> objectMapper.readValue(x, Map.class)).returns(Map.class);
Map<String, String> split = config.getSplit();
partitionByTableAndPrimarykey(mapOperator, tableMap);
return mapOperator.process(new ProcessFunction<Map, Map>() {
@Override
public void processElement(Map map, ProcessFunction<Map, Map>.Context ctx, Collector<Map> out) {
Expand Down Expand Up @@ -238,6 +241,33 @@ public DataStreamSource<String> build(
return dataStreamSource;
}

protected void partitionByTableAndPrimarykey(
SingleOutputStreamOperator<Map> mapOperator, Map<String, Table> tableMap) {
mapOperator.partitionCustom(
new Partitioner<String>() {
@Override
public int partition(String key, int numPartitions) {
return Math.abs(key.hashCode()) % numPartitions;
}
},
map -> {
LinkedHashMap source = (LinkedHashMap) map.get("source");
String tableName = createTableName(source, config.getSchemaFieldName(), config.getSplit());
Table table = tableMap.get(tableName);
List<String> primaryKeys = table.getColumns().stream()
.map(column -> {
if (column.isKeyFlag()) {
return column.getName();
}
return "";
})
.collect(Collectors.toList());

return tableName + String.join("_", primaryKeys);
});
mapOperator.name("PartitionByPrimarykey");
}

protected void executeCatalogStatement(CustomTableEnvironment customTableEnvironment) {}
;
}
22 changes: 22 additions & 0 deletions dinky-core/src/main/java/org/dinky/trans/ddl/CDCSource.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@
package org.dinky.trans.ddl;

import org.dinky.assertion.Asserts;
import org.dinky.data.model.FlinkCDCConfig;
import org.dinky.parser.SingleSqlParserFactory;

import java.util.ArrayList;
Expand Down Expand Up @@ -222,6 +223,27 @@ private static Map<String, String> getKeyValue(List<String> list) {
return map;
}

public FlinkCDCConfig buildFlinkCDCConfig() {
return new FlinkCDCConfig(
connector,
hostname,
port,
username,
password,
checkpoint,
parallelism,
database,
schema,
table,
startupMode,
split,
debezium,
source,
sink,
sinks,
jdbc);
}

public String getConnector() {
return connector;
}
Expand Down
19 changes: 1 addition & 18 deletions dinky-core/src/main/java/org/dinky/trans/ddl/CreateCDCSourceOperation.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -75,24 +75,7 @@ public Operation create(String statement) {
public TableResult execute(Executor executor) {
logger.info("Start build CDCSOURCE Task...");
CDCSource cdcSource = CDCSource.build(statement);
FlinkCDCConfig config = new FlinkCDCConfig(
cdcSource.getConnector(),
cdcSource.getHostname(),
cdcSource.getPort(),
cdcSource.getUsername(),
cdcSource.getPassword(),
cdcSource.getCheckpoint(),
cdcSource.getParallelism(),
cdcSource.getDatabase(),
cdcSource.getSchema(),
cdcSource.getTable(),
cdcSource.getStartupMode(),
cdcSource.getSplit(),
cdcSource.getDebezium(),
cdcSource.getSource(),
cdcSource.getSink(),
cdcSource.getSinks(),
cdcSource.getJdbc());
FlinkCDCConfig config = cdcSource.buildFlinkCDCConfig();
try {
CDCBuilder cdcBuilder = CDCBuilderFactory.buildCDCBuilder(config);
Map<String, Map<String, String>> allConfigMap = cdcBuilder.parseMetaDataConfigs();
Expand Down
Loading