Python 消费Kafka手动提交 批量存入Elasticsearch
一、第三方包选择
pip install kafka，对比了kafka和pykafka，还是选择kafka，消费速度更快
pip install elasticsearch==7.12.0(ES版本)
二、创建es连接对象
from elasticsearch import Elasticsearch
from elasticsearch.helpers import bulkclass Create_ES(object):_instance = Nonedef __new__(cls, *args, **kwargs):if cls._instance is None:cls._instance = super().__new__(cls)return cls._instancedef __init__(self, hosts):try:self.es = Elasticsearch([{'host':host, 'port':9200}])except Exception as e:print('Connect ES Fail db:{} error:{}'.format(hosts, str(e)))def get_conn(self):return self.esdef set_multi_data(self, datas):'''批量插入数据'''success = bulk(self.es, datas, raise_on_error=True)return success三、消费kafka数据
from kafka import KafkaConsumer, TopicPartition, OffsetAndMetadata
from . import Create_ESclass AppKfkConsumer(object):def __init__(self):self.server = 'localhost:9092'self.topic = KAFKA_TOPICself.consumer = Noneself.tp = Noneself.consumer_timeout_ms = 5000 # 设置消费超时时间，self.type = 'members'self.group_id = 'test1' # 设置消费group_id,避免重复消费self.es_index = 'index' # es的indexdef get_connect(self):self.consumer = KafkaConsumer(group_id=self.group_id,auto_offset_reset='earliest', # 从最早的数据开始消费bootstrap_servers=self.server,enable_auto_commit=False, # 关闭自动提交consumer_timeout_ms=self.consumer_timeout_ms)self.tp = TopicPartition(topic=self.topic, partition=0) # 设置我们要消费的分区self.consumer.assign([self.tp]) # 由consumer对象分配分区def beginConsumer(self):now_offset = 0 # 当前偏移量es_conn = Create_ES()Actions = []while True:for message in self.consumer:now_offset = message.offset # 获取当前偏移量data = eval(message.value.decode()) # 解析数据action = {"_index": self.es_index,"_type": self.type,"_source": data}Actions.append(action)if len(Actions) >= 50000:result = es_conn.set_multi_data(Actions) # 批量插入数据Actions = []# 提交偏移量，now_offset+1的原因是因为我发现如果不加1，下次消费会从上次消费最后一条数据开始，重复消费self.consumer.commit(offsets={tp:(OffsetAndMetadata(now_offset+1, None))})if len(Actions) > 0:result = es_conn.set_multi_data(Actions)Actions = []self.consumer.commit(offsets={tp:(OffsetAndMetadata(now_offset+1, None))})def delconnect(self):self.consumer.close()# 执行任务
ks = AppKfkConsumer()
ks.get_connect()
ks.beginConsumer()