kafka重置消费位点java代码实现

package cn.yzw.jc.ppfs.biz.shared.mq;

import lombok.extern.slf4j.Slf4j;
import org.apache.commons.collections4.CollectionUtils;
import org.apache.kafka.clients.consumer.*;
import org.apache.kafka.common.PartitionInfo;
import org.apache.kafka.common.TopicPartition;
import org.apache.kafka.common.serialization.IntegerDeserializer;
import org.apache.kafka.common.serialization.StringDeserializer;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.stereotype.Service;

import java.util.*;

@Service
@Slf4j
public class KafkaConsumerOffsetManager {

    private static String topic = "dba.mssql.jcmalldb.YZ_Mall.Order";

    @Value("${spring.kafka.bootstrap-servers}")
    private String kafkaService;
    @Value("${ppfs.kafka.group}")
    private String groupId;

    /**
     * 重置kafka消费位点
     *
     * @param topicPartitionMap
     */
    public void resetPartitionSites(Map<String, Map<Integer, Long>> topicPartitionMap) {
        Map<String, Object> config = new HashMap<String, Object>();
        //bootserverscon
        config.put(ConsumerConfig.BOOTSTRAP_SERVERS_CONFIG, kafkaService);
        //valuedeserilizer
        config.put(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, IntegerDeserializer.class);
        config.put(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class);
        //groupid
        config.put(ConsumerConfig.GROUP_ID_CONFIG, groupId);
        //如果找不到偏移量，设置earliest,则从最新消费开始,也就是消费者一开始最新消费的时候
        //一定要注意顺序，读取时候的顺序会影响
//        config.put(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, "earliest");
        //消费者
        KafkaConsumer<Integer, String> consumer = new KafkaConsumer<Integer, String>(config);
        for (Map.Entry<String, Map<Integer, Long>> entry : topicPartitionMap.entrySet()) {
            Map<TopicPartition, OffsetAndMetadata> offset = new HashMap<>();
            List<PartitionInfo> partitionInfos = consumer.partitionsFor(entry.getKey());
            if (CollectionUtils.isEmpty(partitionInfos)) {
                log.warn("topic{}未获取到分区", entry.getKey());
                continue;
            }
            if (partitionInfos.size() > 0) {
                for (PartitionInfo p : partitionInfos) {
                    if (entry.getValue().get(p.partition()) == null) {
                        log.warn("topic{}未获取到{}分区位点", entry.getKey(), p.partition());
                        continue;
                    }
                    offset.put(new TopicPartition(entry.getKey(), p.partition()), new OffsetAndMetadata(entry.getValue().get(p.partition())));
                }
            }
            if (!offset.isEmpty()) {
                consumer.commitSync(offset);
            }
        }

    }

    public static void main(String[] args) {
        //配置必要的参数
        //准备一个map集合放置参数
        Map<String, Object> config = new HashMap<String, Object>();
        //bootserverscon
        config.put(ConsumerConfig.BOOTSTRAP_SERVERS_CONFIG, "172.16.1.43:6667,172.16.1.44:6667,172.16.1.45:6667");
        //valuedeserilizer
        config.put(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, IntegerDeserializer.class);
        config.put(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class);
        //groupid
        config.put(ConsumerConfig.GROUP_ID_CONFIG, "ppfs-qa");
        //如果找不到偏移量，设置earliest,则从最新消费开始,也就是消费者一开始最新消费的时候
        //一定要注意顺序，读取时候的顺序会影响
        config.put(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, "earliest");
        //此处是把消费者的偏移量重置到生产者最顶端
        Map<TopicPartition, OffsetAndMetadata> hashMaps = new HashMap<TopicPartition, OffsetAndMetadata>();
        hashMaps.put(new TopicPartition(topic, 0), new OffsetAndMetadata(0));
        //消费者
        KafkaConsumer<Integer, String> consumer = new KafkaConsumer<Integer, String>(config);
        //放置刚刚设置的偏移量
        consumer.commitSync(hashMaps);
        //先订阅后消费
        consumer.subscribe(Arrays.asList(topic));
        // 批量从主题的分区拉取消息
        //final ConsumerRecords<Integer, String> consumerRecords = consumer.poll(3_000);
        ConsumerRecords<Integer, String> consumerRecords = consumer.poll(3000);

        //遍历本次从主题的分区拉取的批量消息
        consumerRecords.forEach(new java.util.function.Consumer<ConsumerRecord<Integer, String>>() {
            @Override
            public void accept(ConsumerRecord<Integer, String> consumerRecord) {
                System.out.println(
                        consumerRecord.topic() + "\t"
                                + consumerRecord.offset() + "\t"
                                + consumerRecord.key() + "\t"
                                + consumerRecord.value() + "\t"
                );
            }
        });
        consumer.close();


    }
}