Question

我试图执行以下查询，并且在处理约1.5小时后中止会被中止：

create volatile table test2 as 
(
select 
    a.*,
    b.client_dk
from test1 as a
inner join  payroll_clients as b 
    on a.payroll_agrmnt_id = b.lbt_lst_payroll_agrmnt_id
)
WITH DATA 
    ON 
COMMIT    PRESERVE ROWS;

表test1包含~1500行，payroll_agrmnt_id为unique。

表payroll_clients包含~151 mln 行（是的），client_dk是唯一的

我做错了什么？

p.s。：已添加＆＃34;显示表格＆＃34;对于两个表：

show table payroll_clients;

CREATE SET VOLATILE TABLE "PAVLENKO-GA".payroll_clients ,NO FALLBACK ,
     CHECKSUM = DEFAULT,
     DEFAULT MERGEBLOCKRATIO,
     LOG
     (
      client_dk BIGINT,
      lbt_lst_payroll_agrmnt_id BIGINT)
PRIMARY INDEX ( client_dk ,lbt_lst_payroll_agrmnt_id )
ON COMMIT PRESERVE ROWS;



show table test1;

CREATE SET VOLATILE TABLE "PAVLENKO-GA".test1 ,NO FALLBACK ,
     CHECKSUM = DEFAULT,
     DEFAULT MERGEBLOCKRATIO,
     LOG
     (
      INN FLOAT,
      DOG_NUM FLOAT,
      payroll_agrmnt_id BIGINT)
PRIMARY INDEX ( payroll_agrmnt_id )
ON COMMIT PRESERVE ROWS;

Answer 1

解决方案由@dnoeth提供 - 应为目标表定义主索引。 THX

Teradata：由于查询不理想而导致会话中止

1 个答案: