ksqldb实现异常交易检测

最新推荐文章于 2024-09-08 07:58:35 发布

ZHANGJNWEI

最新推荐文章于 2024-09-08 07:58:35 发布

阅读量355

点赞数

文章标签： kafka java 大数据

本文链接：https://blog.csdn.net/ZHANGJNWEI/article/details/129691021

版权

本教程展示了如何使用ksqldb创建一个事件驱动的微服务，用于识别可疑活动并通知客户。通过ksqldb检测异常，并使用SendGrid的简单Kafka消费者发送警报邮件。

摘要由CSDN通过智能技术生成

Imagine that you work at a financial services company which clears many credit card transactions each day. You want to prevent malicious activity in your customer base. When a high number of transactions occurs in a narrow window of time, you want to notify the cardholder of suspicious activity.

This tutorial shows how to create an event-driven microservice that identifies suspicious activity and notifies customers. It demonstrates finding anomalies with ksqlDB and sending alert emails using a simple Kafka consumer with SendGrid.

Start the stack¶

To get started, create the following docker-compose.yml file. This specifies all the infrastructure that you'll need to run this tutorial:

---
version: '2'

services:
  zookeeper:
    image: confluentinc/cp-zookeeper:7.3.0
    hostname: zookeeper
    container_name: zookeeper
    ports:
      - "2181:2181"
    environment:
      ZOOKEEPER_CLIENT_PORT: 2181
      ZOOKEEPER_TICK_TIME: 2000

  broker:
    image: confluentinc/cp-kafka:7.3.0
    hostname: broker
    container_name: broker
    depends_on:
      - zookeeper
    ports:
      - "29092:29092"
    environment:
      KAFKA_BROKER_ID: 1
      KAFKA_ZOOKEEPER_CONNECT: 'zookeeper:2181'
      KAFKA_LISTENER_SECURITY_PROTOCOL_MAP: PLAINTEXT:PLAINTEXT,PLAINTEXT_HOST:PLAINTEXT
      KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://broker:9092,PLAINTEXT_HOST://localhost:29092
      KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR: 1
      KAFKA_GROUP_INITIAL_REBALANCE_DELAY_MS: 0
      KAFKA_TRANSACTION_STATE_LOG_MIN_ISR: 1
      KAFKA_TRANSACTION_STATE_LOG_REPLICATION_FACTOR: 1

  schema-registry:
    image: confluentinc/cp-schema-registry:7.3.0
    hostname: schema-registry
    container_name: schema-registry
    depends_on:
      - zookeeper
      - broker
    ports:
      - "8081:8081"
    environment:
      SCHEMA_REGISTRY_HOST_NAME: schema-registry
      SCHEMA_REGISTRY_KAFKASTORE_BOOTSTRAP_SERVERS: "PLAINTEXT://broker:9092"

  ksqldb-server:
    image: confluentinc/ksqldb-server:0.28.2
    hostname: ksqldb-server
    container_name: ksqldb-server
    depends_on:
      - broker
      - schema-registry
    ports:
      - "8088:8088"
    environment:
      KSQL_LISTENERS: "http://0.0.0.0:8088"
      KSQL_BOOTSTRAP_SERVERS: "broker:9092"
      KSQL_KSQL_SCHEMA_REGISTRY_URL: "http://schema-registry:8081"
      KSQL_KSQL_LOGGING_PROCESSING_STREAM_AUTO_CREATE: "true"
      KSQL_KSQL_LOGGING_PROCESSING_TOPIC_AUTO_CREATE: "true"
      # Configuration to embed Kafka Connect support.
      KSQL_CONNECT_GROUP_ID: "ksql-connect-cluster"
      KSQL_CONNECT_BOOTSTRAP_SERVERS: "broker:9092"
      KSQL_CONNECT_KEY_CONVERTER: "org.apache.kafka.connect.storage.StringConverter"
      KSQL_CONNECT_VALUE_CONVERTER: "io.confluent.connect.avro.AvroConverter"
      KSQL_CONNECT_VALUE_CONVERTER_SCHEMA_REGISTRY_URL: "http://schema-registry:8081"
      KSQL_CONNECT_CONFIG_STORAGE_TOPIC: "_ksql-connect-configs"
      KSQL_CONNECT_OFFSET_STORAGE_TOPIC: "_ksql-connect-offsets"
      KSQL_CONNECT_STATUS_STORAGE_TOPIC: "_ksql-connect-statuses"
      KSQL_CONNECT_CONFIG_STORAGE_REPLICATION_FACTOR: 1
      KSQL_CONNECT_OFFSET_STORAGE_REPLICATION_FACTOR: 1
      KSQL_CONNECT_STATUS_STORAGE_REPLICATION_FACTOR: 1
      KSQL_CONNECT_PLUGIN_PATH: "/usr/share/kafka/plugins"

  ksqldb-cli:
    image: confluentinc/ksqldb-cli:0.28.2
    container_name: ksqldb-cli
    depends_on:
      - broker
      - ksqldb-server
    entrypoint: /bin/sh
    tty: true

Bring up the stack by running:

1	`docker-compose up`

Create the transactions stream¶

Connect to ksqlDB's server by using its interactive CLI. Run the following command from your host:

1	`docker exec -it ksqldb-cli ksql http://ksqldb-server:8088`

Before you issue more commands, tell ksqlDB to start all queries from earliest point in each topic:

1	`SET 'auto.offset.reset' = 'earliest';`

We want to model a stream of credit card transactions from which we'll look for anomalous activity. To do that, create a ksqlDB stream to represent the transactions. Each transaction has a few key pieces of information, like the card number, amount, and email address that it's associated with. Because the specified topic (transactions) does not exist yet, ksqlDB creates it on your behalf.

CREATE STREAM transactions (
    tx_id VARCHAR KEY,
    email_address VARCHAR,
    card_number VARCHAR,
    timestamp VARCHAR,
    amount DECIMAL(12, 2)
) WITH (
    kafka_topic = 'transactions',
    partitions = 8,
    value_format = 'avro',
    timestamp = 'timestamp',
    timestamp_format = 'yyyy-MM-dd''T''HH:mm:ss'
);

Notice that this stream is configured with a custom timestamp to signal that event-time should be used instead of processing-time. What this means is that when ksqlDB does time-related operations over the stream, it uses the timestamp column to measure time, not the current time of the operating system. This makes it possible to handle out-of-order events.

The stream is also configured to use the Avro format for the value part of the underlying Kafka records