前言
有一位同學(xué)正在調(diào)研MLSQL Stack對流的支持。然后說了流調(diào)試其實挺困難的。經(jīng)過實踐，希望實現(xiàn)如下三點：
- 能隨時查看最新固定條數(shù)的Kafka數(shù)據(jù)
- 調(diào)試結(jié)果（sink）能打印在web控制臺
- 流程序能自動推測json schema(現(xiàn)在spark是不行的)
實現(xiàn)這三個點之后，我發(fā)現(xiàn)調(diào)試確實就變得簡單很多了。
流程
首先我新建了一個kaf_write.mlsql,里面方便我往Kafka里寫數(shù)據(jù)：
set abc='''
{ "x": 100, "y": 200, "z": 200 ,"dataType":"A group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
{ "x": 120, "y": 100, "z": 260 ,"dataType":"B group"}
''';
load jsonStr.`abc` as table1;
select to_json(struct(*)) as value from table1 as table2;
save append table2 as kafka.`wow` where
kafka.bootstrap.servers="127.0.0.1:9092";
這樣我每次運行，數(shù)據(jù)就能寫入到Kafka.
接著，我寫完后，需要看看數(shù)據(jù)是不是真的都寫進去了，寫成了什么樣子：
!kafkaTool sampleData 10 records from "127.0.0.1:9092" wow;
這句話表示，我要采樣Kafka 10條Kafka數(shù)據(jù)，該Kafka的地址為127.0.0.1:9092,主題為wow.運行結(jié)果如下：
沒有什么問題。接著我寫一個非常簡單的流式程序：
-- the stream name, should be uniq.
set streamName="streamExample";
-- use kafkaTool to infer schema from kafka
!kafkaTool registerSchema 2 records from "127.0.0.1:9092" wow;
load kafka.`wow` options
kafka.bootstrap.servers="127.0.0.1:9092"
as newkafkatable1;
select * from newkafkatable1
as table21;
-- print in webConsole instead of terminal console.
save append table21
as webConsole.``
options mode="Append"
and duration="15"
and checkpointLocation="/tmp/s-cpl4";
運行結(jié)果如下：
在終端我們也可以看到實時效果了。
補充
當(dāng)然，MLSQL Stack 還有對流還有兩個特別好地方，第一個是你可以對流的事件設(shè)置http協(xié)議的callback，以及對流的處理結(jié)果再使用批SQL進行處理，最后入庫。參看如下腳本：
-- the stream name, should be uniq.
set streamName="streamExample";
-- mock some data.
set data='''
{"key":"yes","value":"no","topic":"test","partition":0,"offset":0,"timestamp":"2008-01-24 18:01:01.001","timestampType":0}
{"key":"yes","value":"no","topic":"test","partition":0,"offset":1,"timestamp":"2008-01-24 18:01:01.002","timestampType":0}
{"key":"yes","value":"no","topic":"test","partition":0,"offset":2,"timestamp":"2008-01-24 18:01:01.003","timestampType":0}
{"key":"yes","value":"no","topic":"test","partition":0,"offset":3,"timestamp":"2008-01-24 18:01:01.003","timestampType":0}
{"key":"yes","value":"no","topic":"test","partition":0,"offset":4,"timestamp":"2008-01-24 18:01:01.003","timestampType":0}
{"key":"yes","value":"no","topic":"test","partition":0,"offset":5,"timestamp":"2008-01-24 18:01:01.003","timestampType":0}
''';
-- load data as table
load jsonStr.`data` as datasource;
-- convert table as stream source
load mockStream.`datasource` options
stepSizeRange="0-3"
as newkafkatable1;
-- aggregation
select cast(value as string) as k from newkafkatable1
as table21;
!callback post "http://127.0.0.1:9002/api_v1/test" when "started,progress,terminated";
-- output the the result to console.
save append table21
as custom.``
options mode="append"
and duration="15"
and sourceTable="jack"
and code='''
select count(*) as c from jack as newjack;
save append newjack as parquet.`/tmp/jack`;
'''
and checkpointLocation="/tmp/cpl15";
總結(jié)
以上就是這篇文章的全部內(nèi)容了，希望本文的內(nèi)容對大家的學(xué)習(xí)或者工作具有一定的參考學(xué)習(xí)價值，謝謝大家對腳本之家的支持。
您可能感興趣的文章:- Mysql LONGBLOB 類型存儲二進制數(shù)據(jù) (修改+調(diào)試+整理)
- Mysql LONGTEXT 類型存儲大文件(二進制也可以) (修改+調(diào)試+整理)
- Mysql 插入中文及中文查詢 (修改+調(diào)試)
- 新手配置 PHP 調(diào)試環(huán)境（IIS+PHP+MYSQL）
- MySQL UDF調(diào)試方式debugview的相關(guān)方法
- 分享101個MySQL調(diào)試與優(yōu)化技巧
- GDB調(diào)試Mysql實戰(zhàn)之源碼編譯安裝
