kafka删除topic消息的四种方式[通俗易懂]

#日志清理策略选择有：delete和compact主要针对过期数据的处理，或是日志文件达到限制的额度，会被 topic创建时的指定参数覆盖
log.cleanup.policy = delete

# 注意:下面有两种配置，一种是基于时间的策略，另种是基于日志文件大小的策略，两种策略同是配置的话，只要满足其中种策略，则触发Log删除的操作。删除操作总是先删除最旧的日志
# 消息在Kafka中保存的时间，168小时之前的1og， 可以被删除掉，根据policy处理数据。
log.retention.hours=4

# 当剩余空间低于log.retention.bytes字节，则开始删除1og
log.retention.bytes=37580963840

# 每隔300000ms, logcleaner线程将检查一次，看是否符合上述保留策略的消息可以被删除
log.retention.check.interval.ms=1000

package com.censoft.kafkaAdmin;
import org.apache.kafka.clients.admin.*;
import org.apache.kafka.common.KafkaFuture;
import org.apache.kafka.common.TopicPartition;
import java.sql.*;
import java.util.HashMap;
import java.util.Map;
import java.util.Properties;
import java.util.concurrent.ExecutionException;
/**
* @author zy Zhang
* @version : 1.0
* @Description
* @since 2020/7/13 16:02
*/
public class DeleteReordsByOffset {
public static void main(String[] args) throws ClassNotFoundException {
// 1.创建kafkaAdminClient
Properties properties = new Properties();
properties.put("bootstrap.servers","192.168.27.111:9092");
AdminClient kafkaAdminClient = KafkaAdminClient.create(properties);
// 2.从数据库获取需要删除的消息
Class.forName("com.mysql.jdbc.Driver");
Map<TopicPartition, RecordsToDelete> recordsToDelete = new HashMap<>();
String url  = "jdbc:mysql://localhost:3306/test?useSSL=false&amp;useUnicode=true&amp;characterEncoding=UTF-8";
String user = "root";
String password = "123456";
Connection conn = null;
Statement statement = null;
ResultSet res = null;
String sql = "SELECT Topic, KafkaPartition, UntilOffset FROM Kafka_Offset;";
try {
conn = DriverManager.getConnection(url, user, password);
statement = conn.createStatement();
res = statement.executeQuery(sql);
if (res != null) {
while (res.next()) {
String topic = res.getString("Topic");
Integer partition = res.getInt("KafkaPartition");
Long offset = res.getLong("UntilOffset");
TopicPartition topicPartition = new TopicPartition(topic, partition);
RecordsToDelete recordsToDelete1 = RecordsToDelete.beforeOffset(offset);
recordsToDelete.put(topicPartition, recordsToDelete1);
}
}
} catch (SQLException e) {
e.printStackTrace();
} finally {
if (statement != null) {
try {
statement.close();
} catch (SQLException e) {
e.printStackTrace();
}
}
if (conn != null) {
try {
conn.close();
} catch (SQLException e) {
e.printStackTrace();
}
}
}
// 3.执行删除
DeleteRecordsResult result = kafkaAdminClient.deleteRecords(recordsToDelete);
Map<TopicPartition, KafkaFuture<DeletedRecords>> lowWatermarks = result.lowWatermarks();
try {
for (Map.Entry<TopicPartition, KafkaFuture<DeletedRecords>> entry : lowWatermarks.entrySet()) {
System.out.println(entry.getKey().topic() + " " + entry.getKey().partition() + " " + entry.getValue().get().lowWatermark());
}
} catch (InterruptedException | ExecutionException e) {
e.printStackTrace();
}
kafkaAdminClient.close();
}
}

		目前发现通过这种方法起到的效果是：
topic的起始偏移量会被定位到传入的recordsToDelete指定的位置
但是并没有将磁盘中存储的数据删除
如果我找到在磁盘删除的方法会继续更新，看下面

eg: test-0下有三个log文件
00000000000000000000.log， 00000000000000000036.log， 00000000000000000136.log
我们修改起始偏移量=100
那么deleteLogStartOffsetBreachedSegments运行时会判定00000000000000000000.log是可以删除的