Flink SQL-Client 的使用
flink sql client 介绍
The SQL Client aims to provide an easy way of writing, debugging, and submitting table programs to a Flink cluster without a single line of Java or Scala code. The SQL Client CLI allows for retrieving and visualizing real-time results from the running distributed application on the command line.
flink sql-client 是一种实用的工具，方便 flink 开发人员编写，调试，提交实时table代码, 不用编写 Java 或 Scala代码。同时在 sql-client 上能够可视化的看到实时统计的 retract 和 append 结果。
部署环境 (单机)
- java 1.8
- zookeeper 3.4.13
- kafka 0.11
- flink 1.6
- 启动
zookeeper
yizhou@pro:~$ zkServer start
ZooKeeper JMX enabled by default
Using config: /usr/local/etc/zookeeper/zoo.cfg
Starting zookeeper ... STARTED
yizhou@pro:~$ zkServer status
ZooKeeper JMX enabled by default
Using config: /usr/local/etc/zookeeper/zoo.cfg
Mode: standalone
- 启动
kafka
yizhou@pro:${KAFKA_HOME}$ nohup bin/kafka-server-start.sh config/server.properties &
[1] 70358
# 查看 topic
yizhou@pro:${KAFKA_HOME}$ bin/kafka-topics.sh --list --zookeeper localhost:2181
__consumer_offsets
idea
order_sql
# 往 order_sql 这个 topic发送 json 消息
yizhou@pro:${KAFKA_HOME}$ bin/kafka-console-producer.sh --broker-list localhost:9092 --topic order_sql
>{"order_id": "1","shop_id": "AF18","member_id": "3410211","trade_amt": "100.00","pay_time": "1556420980000"}
>{"order_id": "2","shop_id": "AF20","member_id": "3410213","trade_amt": "130.00","pay_time": "1556421040000"}
>{"order_id": "3","shop_id": "AF18","member_id": "3410212","trade_amt": "120.00","pay_time": "1556421100000"}
>{"order_id": "4","shop_id": "AF19","member_id": "3410212","trade_amt": "100.00","pay_time": "1556421120000"}
>{"order_id": "5","shop_id": "AF18","member_id": "3410211","trade_amt": "150.00","pay_time": "1556421480000"}
>{"order_id": "6","shop_id": "AF18","member_id": "3410211","trade_amt": "110.00","pay_time": "1556421510000"}
>{"order_id": "7","shop_id": "AF19","member_id": "3410213","trade_amt": "110.00","pay_time": "1556421570000"}
>{"order_id": "8","shop_id": "AF20","member_id": "3410211","trade_amt": "100.00","pay_time": "1556421630000"}
>{"order_id": "9","shop_id": "AF17","member_id": "3410212","trade_amt": "110.00","pay_time": "1556421655000"}
>
- 启动
flink
yizhou@pro:${FLINK_HOME}/bin$ ./start-cluster.sh
Starting cluster.
Starting standalonesession daemon on host pro.local.
Starting taskexecutor daemon on host pro.local.
注册 kafka 消息为 flink sql 的动态表
将kafka消息映射为flink sql 的动态表，是非常重要的操作。整个操作通过配置 ${FLINK_HOME}/conf 目录下的 yaml 文件实现。以下例子是将 kafka 的 topic: order_sql, 映射为 table。
#==============================================================================
# Table Sources
#==============================================================================
# Define table sources and sinks here.
tables: # empty list
# A typical table source definition looks like:
- name: orders
type: source
update-mode: append
connector:
property-version: 1
type: kafka
version: 0.11
topic: order_sql
startup-mode: earliest-offset
properties:
- key: zookeeper.connect
value: localhost:2181
- key: bootstrap.servers
value: localhost:9092
- key: group.id
value: test-consumer-group
format:
property-version: 1
type: json
schema: "ROW(order_id LONG, shop_id VARCHAR, member_id LONG, trade_amt DOUBLE, pay_time TIMESTAMP)"
schema:
- name: order_id
type: LONG
- name: shop_id
type: VARCHAR
- name: member_id
type: LONG
- name: trade_amt
type: DOUBLE
- name: payment_time
type: TIMESTAMP
rowtime:
timestamps:
type: "from-field"
from: "pay_time"
watermarks:
type: "periodic-bounded"
delay: "60000"
#==============================================================================
# User-defined functions
#==============================================================================
# Define scalar, aggregate, or table functions here.
functions: [] # empty list
# A typical function definition looks like:
# - name: ...
# from: class
# class: ...
# constructor: ...
#==============================================================================
# Execution properties
#==============================================================================
# Execution properties allow for changing the behavior of a table program.
execution:
# 'batch' or 'streaming' execution
type: streaming
# allow 'event-time' or only 'processing-time' in sources
time-characteristic: event-time
# interval in ms for emitting periodic watermarks
periodic-watermarks-interval: 200
# 'changelog' or 'table' presentation of results
result-mode: table
# maximum number of maintained rows in 'table' presentation of results
max-table-result-rows: 1000000
# parallelism of the program
parallelism: 1
# maximum parallelism
max-parallelism: 128
# minimum idle state retention in ms
min-idle-state-retention: 3600000
# maximum idle state retention in ms
max-idle-state-retention: 7200000
#==============================================================================
# Deployment properties
#==============================================================================
# Deployment properties allow for describing the cluster to which table
# programs are submitted to.
deployment:
# general cluster communication timeout in ms
response-timeout: 5000
# (optional) address from cluster to gateway
gateway-address: ""
# (optional) port from cluster to gateway
gateway-port: 0
从配置文件中可以看出，flink 中注册的表名为 orders，作为source的数据源表，append的方式不断添加。以 0.11 版本的 kafka 的作为 connector, topic 信息、消息偏移量、zookeeper 地址、broker 地址、消费组信息都写入配置。同时 kafka 消息的 json 消息如何映射到 flink sql table 的 shecma，选择pay_time 作为 event-time，watermark 设置为60s。
启动 flink sql client
命令行启动 bin/sql-client.sh embedded -d conf/sql.my.yaml -l sql-libs/。 其中 sql.my.yaml 是上述的 yaml 配置文件名称。${FLINK_HOME}/sql-libs 目录下需要提前下载 flink-connector-kafka-0.11 和 flink-json-1.6.1-sql-jar.jar 两个jar包
yizhou@pro:${FLINK_HOME}$ bin/sql-client.sh embedded -d conf/sql.my.yaml -l sql-libs/
Reading default environment from: file:/usr/local/Cellar/apache-flink/1.6.2/libexec/conf/sql.my.yaml
No session environment specified.
▒▓██▓██▒
▓████▒▒█▓▒▓███▓▒
▓███▓░░ ▒▒▒▓██▒ ▒
░██▒ ▒▒▓▓█▓▓▒░ ▒████
██▒ ░▒▓███▒ ▒█▒█▒
░▓█ ███ ▓░▒██
▓█ ▒▒▒▒▒▓██▓░▒░▓▓█
█░ █ ▒▒░ ███▓▓█ ▒█▒▒▒
████░ ▒▓█▓ ██▒▒▒ ▓███▒
░▒█▓▓██ ▓█▒ ▓█▒▓██▓ ░█░
▓░▒▓████▒ ██ ▒█ █▓░▒█▒░▒█▒
███▓░██▓ ▓█ █ █▓ ▒▓█▓▓█▒
░██▓ ░█░ █ █▒ ▒█████▓▒ ██▓░▒
███░ ░ █░ ▓ ░█ █████▒░░ ░█░▓ ▓░
██▓█ ▒▒▓▒ ▓███████▓░ ▒█▒ ▒▓ ▓██▓
▒██▓ ▓█ █▓█ ░▒█████▓▓▒░ ██▒▒ █ ▒ ▓█▒
▓█▓ ▓█ ██▓ ░▓▓▓▓▓▓▓▒ ▒██▓ ░█▒
▓█ █ ▓███▓▒░ ░▓▓▓███▓ ░▒░ ▓█
██▓ ██▒ ░▒▓▓███▓▓▓▓▓██████▓▒ ▓███ █
▓███▒ ███ ░▓▓▒░░ ░▓████▓░ ░▒▓▒ █▓
█▓▒▒▓▓██ ░▒▒░░░▒▒▒▒▓██▓░ █▓
██ ▓░▒█ ▓▓▓▓▒░░ ▒█▓ ▒▓▓██▓ ▓▒ ▒▒▓
▓█▓ ▓▒█ █▓░ ░▒▓▓██▒ ░▓█▒ ▒▒▒░▒▒▓█████▒
██░ ▓█▒█▒ ▒▓▓▒ ▓█ █░ ░░░░ ░█▒
▓█ ▒█▓ ░ █░ ▒█ █▓
█▓ ██ █░ ▓▓ ▒█▓▓▓▒█░
█▓ ░▓██░ ▓▒ ▓█▓▒░░░▒▓█░ ▒█
██ ▓█▓░ ▒ ░▒█▒██▒ ▓▓
▓█▒ ▒█▓▒░ ▒▒ █▒█▓▒▒░░▒██
░██▒ ▒▓▓▒ ▓██▓▒█▒ ░▓▓▓▓▒█▓
░▓██▒ ▓░ ▒█▓█ ░░▒▒▒
▒▓▓▓▓▓▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒░░▓▓ ▓░▒█░
______ _ _ _ _____ ____ _ _____ _ _ _ BETA
| ____| (_) | | / ____|/ __ \| | / ____| (_) | |
| |__ | |_ _ __ | | __ | (___ | | | | | | | | |_ ___ _ __ | |_
| __| | | | '_ \| |/ / \___ \| | | | | | | | | |/ _ \ '_ \| __|
| | | | | | | | < ____) | |__| | |____ | |____| | | __/ | | | |_
|_| |_|_|_| |_|_|\_\ |_____/ \___\_\______| \_____|_|_|\___|_| |_|\__|
Welcome! Enter HELP to list all available commands. QUIT to exit.
Flink SQL> show tables;
orders
Flink SQL> describe orders;
root
|-- order_id: Long
|-- shop_id: String
|-- member_id: Long
|-- trade_amt: Double
|-- payment_time: TimeIndicatorTypeInfo(rowtime)
Flink SQL>
运行 sql 语句
- 首先执行最简单的
select *
Flink SQL> select * from orders;
select * from orders
- 1分钟固定窗口计算
SELECT
shop_id
, TUMBLE_START(payment_time, INTERVAL '1' MINUTE) AS tumble_start
, TUMBLE_END(payment_time, INTERVAL '1' MINUTE) AS tumble_end
, sum(trade_amt) AS amt
FROM orders
GROUP BY shop_id, TUMBLE(payment_time, INTERVAL '1' MINUTE);
1分钟固定窗口计算
参考资料
- SQL Client
- Connect to External Systems