数据加载

load  excel

load excel.`tmp/upload/exceltest.xls` 
where header="true" 
as hello_world;

load csv

load csv.`tmp/upload/taxi_tour_table_train_simple.csv` where delimiter=","
and header = "true"
as taxi_tour_table_train_simple;

load holo table

load jdbc.`db_2.ads` where directQuery='''
SELECT * FROM ads.inner_t224_car_sl_repurchase_order_detail_a_d limit 10
''' as newtable;

select * from newtable as output;
 

Connection

链接 hologres

SET user="JYIlS*******qXe";
SET password="M1Vpw3YBR******cX8CpxZ9JD";
CONNECT jdbc WHERE
 url="jdbc:postgresql://holo-cn-fb4t8dpbb942-cn-baoding-gwmcloud-d01-internal.hologres.ops.cloud.gwm.cn:80/sl_data_prod"
 and driver="org.postgresql.Driver"
 and user="${user}"
 and password="${password}"
 AS db_2;

如果链接holo后，load数据时，在表的前面增加数据库实例，会报下面错误：

 JDBC direct query should not allow using db prefix. Please just use table

解决办法：1.将数据表放在public中

2.修改byzer引擎配置文件：byzer.properties.override 

增加

cd conf

spark.mlsql.directquery.auth.enable=false

##然后重启

cd bin 

nohup ./byzer.sh restart