0

以下是我为 kafka 生产者编写的 python 编码，我不确定消息是否能够发布到 Kafka Broker。因为消费者端没有收到任何消息。当我使用生产者控制台命令对其进行测试时，我的消费者 python 程序运行良好。

from __future__ import print_function

import sys
from pyspark import SparkContext
from kafka import KafkaClient, SimpleProducer

if __name__ == "__main__":

if len(sys.argv) != 2:
    print("Usage:spark-submit producer1.py <input file>", file=sys.stderr)
    exit(-1)

sc = SparkContext(appName="PythonRegression")

def sendkafka(messages):
    ## Set broker port
    kafka = KafkaClient("localhost:9092")
    producer = SimpleProducer(kafka, async=True, batch_send_every_n=5,  
batch_send_every_t=10)
    send_counts = 0
    for message in messages:
        try:
            print(message)
            ## Set topic name and push messages to the Kafka Broker
            yield producer.send_messages('test', message.encode('utf-8'))
        except Exception, e:
            print("Error: %s" % str(e))
        else:
            send_counts += 1
    print("The count of prediction results which were sent IN THIS PARTITION 
is %d.\n" % send_counts)

## Connect and read the file.    
rawData = sc.textFile(sys.argv[1])

## Find and skip the first row
dataHeader = rawData.first()
data =  rawData.filter(lambda x: x != dataHeader)

## Collect the RDDs.
sentRDD = data.mapPartitions(sendkafka) 
sentRDD.collect()

## Stop file connection
sc.stop()

这是我的“消费者”python 编码

from __future__ import print_function
import sys
from pyspark import SparkContext
from pyspark.streaming import StreamingContext
from pyspark.streaming.kafka import KafkaUtils

if len(sys.argv) < 3:
print ("Program to pulls the messages from kafka brokers.")
print("Usage: consume.py <zk> <topic>", file=sys.stderr)

else:
## Flow
## Loads settings from system properties, for launching of spark-submit.
sc = SparkContext(appName="PythonStreamingKafkaWordCount")

## Create a StreamingContext using an existing SparkContext.
ssc = StreamingContext(sc, 10)

## Get everything after the python script name
zkQuorum, topic = sys.argv[1:]

## Create an input stream that pulls messages from Kafka Brokers.
kvs = KafkaUtils.createStream(ssc, zkQuorum, "spark-streaming-consumer", 
{topic: 1})

## 
lines = kvs.map(lambda x: x[1])

## Print the messages pulled from Kakfa Brokers
lines.pprint()

## Save the pulled messages as file
## lines.saveAsTextFiles("OutputA")

## Start receiving data and processing it
ssc.start()

## Allows the current process to wait for the termination of the context 
ssc.awaitTermination()
4

3 回答 3

0

我通常使用 kafka-console-consumer（Apache Kafka 的一部分）调试此类问题，以从您尝试生成的主题中消费。如果控制台消费者收到消息，您就知道它们到达了 Kafka。

如果你先运行生产者，让它完成，然后启动消费者，那么问题可能是消费者从日志末尾开始，正在等待额外的消息。要么确保您首先启动消费者，要么将其配置为从头开始自动启动（抱歉，不确定如何使用 Python 客户端执行此操作）。

于 2016-09-14T04:58:27.910 回答
0

您可以检查主题中的消息数量是否随着 Produce 请求的增加而增加：

./bin/kafka-run-class.sh kafka.tools.GetOffsetShell \
--broker-list <Kafka_broker_hostname>:<broker_port> --topic Que1 \ 
--time -1 --offsets 1 | awk -F  ":" '{sum += $3} END {print sum}'

如果消息数量在增加，则意味着 Producer 工作正常。

于 2016-09-14T08:02:41.803 回答
0

好吧，我认为我本地的 Zookeeper 或 Kafka 有问题，因为我在另一台服务器上对其进行了测试，它运行良好。但是，感谢那些回复我的人；）

于 2016-09-21T02:54:32.833 回答