大数据平台测试-python往kafka发送数据（脚本开发实战）

水深无声 2024-04-08 09:14 61阅读 0赞

最近参与新项目，kafka鉴权相关的。需要做性能测试，在这里需要大量的数据，因此需要写个脚本。在这里记录下。

连接kafka tools，非常简单，填写下面的地址和改一下名称就可以连接了

![9949f48b8c8648dd8b9f4aa20fd15f54.png][]

这里我们先用kafka tools工具创建一个topic

![f56928a048604209b1f0683a4c28d9df.png][]

分区数和副本数填一下，这里我创建的topic为 ：test\_lqj

![43925060d4e547fe8b46336066e1c638.png][]

这里没看到数据

![1c5c0b03f5ca4823bd7ade73694f6d2e.png][]

编写代码

![6cb28913f1c640ada63d625f0cd10a4d.png][]

发送10个数据测试下

![a73c366cdc1b45dfa8487f7fa9174445.png][]

用kafka tools工具看下

![d3abf4591f30448aa270912d873c6fa1.png][]

成功。

付源码：

# -*- coding: utf-8 -*-
    # @Author  : Liqiju
    # @Time    : 2022/8/16 15:24
    # @File    : send_kafka.py
    # @Software: PyCharm
    
    from kafka import KafkaProducer
    import json
    import random
    import datetime
    import time
    
    
    class Kafka(object):
    
        def __init__(self,topic,json_data,bootstrap_servers):
            self.topic = topic
            self.json_data = json_data
            self.bootstrap_servers = bootstrap_servers
    
        def send_kafka(self):
            byte_data = json.dumps(self.json_data)
            msg = byte_data.encode('utf-8')
            producer.send(self.topic, msg)
    
    if __name__ == '__main__':
        start_time = time.time()
        # topic，这里改为自己的topic
        topic = 'test_lqj'
        # 集群地址，这里改为自己的集群地址，就是Zookeeper Host 地址
        bootstrap_servers = 'XX.XX.XXX.XX,XX.XX.XXX.XX,10.XX.XXX.23'
        # 需要运行的次数，改为自己需要的次数
        sum = 10
        producer = KafkaProducer(bootstrap_servers=bootstrap_servers)
        for i in range(sum):
            score = random.randint(0, 100)  # 1到100随机数
            json_data = {"id": i, "score": score}  # data
            p1 = Kafka(topic, json_data, bootstrap_servers)
            p1.send_kafka()
            print('第'+str(i+1)+'个数据发送成功', datetime.datetime.now())
        producer.close()
        end_time = time.time()
        print('共发送'+ str(sum) + '个数据，'+'共耗时'+ str(end_time-start_time)+"秒")

**以上代码写入JSON含有中文，在kafka tools工具里面查询会有问题：代码改如下**

# -*- coding: utf-8 -*-
    # @Author  : Liqiju
    # @Time    : 2022/8/16 15:24
    # @File    : send_kafka.py
    # @Software: PyCharm
    
    from kafka import KafkaProducer
    import json
    import random
    import datetime
    import time
    
    
    class Kafka(object):
    
        def __init__(self,topic,json_data,bootstrap_servers):
            self.topic = topic
            self.json_data = json_data
            self.bootstrap_servers = bootstrap_servers
    
        def send_kafka(self):
            byte_data = json.dumps(self.json_data)
            msg = byte_data.encode('utf-8')
            producer.send(self.topic, msg)
    
    if __name__ == '__main__':
        start_time = time.time()
        # topic，这里改为自己的topic
        topic = 'liqiju_test_202312_100000000'
        # 集群地址，这里改为自己的集群地址，就是Zookeeper Host 地址
        bootstrap_servers = 'XXXXX' # UAT大数据kafka集群
        # 需要运行的次数，改为自己需要的次数
        sum = 10
        producer = KafkaProducer(bootstrap_servers=bootstrap_servers)
        for i in range(sum):
            json_data = {"id": "apple banner test", "score": "测试"}  # data
            p1 = Kafka(topic, json_data, bootstrap_servers)
            p1.send_kafka()
            print('第'+str(i+1)+'个数据发送成功', datetime.datetime.now())
        producer.close()
        end_time = time.time()
        print('共发送'+ str(sum) + '个数据，'+'共耗时'+ str(end_time-start_time)+"秒")

**跟开发讨论下，开发叫我这样实现，直接上代码**

![7591a5d022f34dd2a8c60a0fc1e181eb.png][]

import threading
    import time
    import datetime
    import json
    from kafka import KafkaProducer
    
    class Kafka(object):
        def __init__(self, topic, json_data, bootstrap_servers):
            self.topic = topic
            self.json_data = json_data
            self.bootstrap_servers = bootstrap_servers
    
        def send_kafka(self):
            byte_data = json.dumps(self.json_data, ensure_ascii=False).encode('utf-8')
            producer.send(self.topic, value=byte_data)
    
    def push_data(thread_id):
        count = 0
        while count < total_data:
            start = count
            end = min(count + batch_size, total_data)
            for i in range(start, end):
                json_data = { "id": "apple banner test", "score": "测试" }  # data
                p1 = Kafka(topic, json_data, bootstrap_servers)
                p1.send_kafka()
                print('线程{} - 第{}个数据发送成功 - {}'.format(thread_id, i+1, datetime.datetime.now()))
                count += 1
                if count >= total_data:
                    break
            time.sleep(8)
    
    if __name__ == '__main__':
        start_time = time.time()
        topic = 'liqiju_quto_test1'
        bootstrap_servers = '' # UAT演示集群
        total_data = 1000000
        thread_count = 2
        batch_size = 8000
        producer = KafkaProducer(bootstrap_servers=bootstrap_servers)
    
        threads = []
        for i in range(thread_count):
            thread = threading.Thread(target=push_data, args=(i+1,))
            threads.append(thread)
            thread.start()
    
        for thread in threads:
            thread.join()
    
        producer.close()
        end_time = time.time()
        print('共发送{}次数据，共耗时{}秒'.format(total_data, end_time-start_time))

以上代码实现开2个线程往topic里面写数据，每个线程推送8000次数据暂停8秒，共推100w

[9949f48b8c8648dd8b9f4aa20fd15f54.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/04/08/e010090a1baa41f59ebf0cb770f2a978.png
[f56928a048604209b1f0683a4c28d9df.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/04/08/ebc511c9ab3a4febb1c3b9df4b85ffda.png
[43925060d4e547fe8b46336066e1c638.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/04/08/98d861420ed940468cbbb04a6430f7ee.png
[1c5c0b03f5ca4823bd7ade73694f6d2e.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/04/08/2306e61b34c640d0bea8d68d6f638a53.png
[6cb28913f1c640ada63d625f0cd10a4d.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/04/08/ed6c4139f6f94fde8fcc424239d6d3d4.png
[a73c366cdc1b45dfa8487f7fa9174445.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/04/08/359f1555ce60469eaa321f8b921d67c8.png
[d3abf4591f30448aa270912d873c6fa1.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/04/08/1e3e1513ca7f4c7da45218c843bb9990.png
[7591a5d022f34dd2a8c60a0fc1e181eb.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/04/08/71d9b43fde6642898adb4da594d9962e.png