lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

[HUDI-2487] Fix JsonKafkaSource cannot filter empty messages from kafka (#3715)

Browse Source
This commit is contained in:
qianchutao
2021-09-28 13:47:15 +08:00
committed by GitHub
parent 36be287121
commit 9067657a5f
2 changed files with 27 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
hudi-utilities/src/main/java/org/apache/hudi/utilities/sources/JsonKafkaSource.java
View File

@@ -69,7 +69,11 @@ public class JsonKafkaSource extends JsonSource {
private JavaRDD<String> toRDD(OffsetRange[] offsetRanges) {
return KafkaUtils.createRDD(sparkContext, offsetGen.getKafkaParams(), offsetRanges,
LocationStrategies.PreferConsistent()).map(x -> (String) x.value());
LocationStrategies.PreferConsistent()).filter(x -> {
String msgValue = (String) x.value();
//Filter null messages from Kafka to prevent Exceptions
return msgValue != null;
}).map(x -> (String) x.value());
}
@Override

22
hudi-utilities/src/test/java/org/apache/hudi/utilities/sources/TestJsonKafkaSource.java
View File

@@ -151,6 +151,28 @@ public class TestJsonKafkaSource extends UtilitiesTestBase {
assertEquals(Option.empty(), fetch4AsRows.getBatch());
}
// test whether empty messages can be filtered
@Test
public void testJsonKafkaSourceFilterNullMsg() {
// topic setup.
testUtils.createTopic(TEST_TOPIC_NAME, 2);
HoodieTestDataGenerator dataGenerator = new HoodieTestDataGenerator();
TypedProperties props = createPropsForJsonSource(null, "earliest");
Source jsonSource = new JsonKafkaSource(props, jsc, sparkSession, schemaProvider, metrics);
SourceFormatAdapter kafkaSource = new SourceFormatAdapter(jsonSource);
// 1. Extract without any checkpoint => get all the data, respecting sourceLimit
assertEquals(Option.empty(), kafkaSource.fetchNewDataInAvroFormat(Option.empty(), Long.MAX_VALUE).getBatch());
// Send 1000 non-null messages to Kafka
testUtils.sendMessages(TEST_TOPIC_NAME, Helpers.jsonifyRecords(dataGenerator.generateInserts("000", 1000)));
// Send 100 null messages to Kafka
testUtils.sendMessages(TEST_TOPIC_NAME,new String[100]);
InputBatch<JavaRDD<GenericRecord>> fetch1 = kafkaSource.fetchNewDataInAvroFormat(Option.empty(), Long.MAX_VALUE);
// Verify that messages with null values are filtered
assertEquals(1000, fetch1.getBatch().get().count());
}
// test case with kafka offset reset strategy
@Test
public void testJsonKafkaSourceResetStrategy() {