SQL 语句运行缓慢 - 有啥想法吗？

Posted 2023-03-24

【中文标题】SQL 语句运行缓慢 - 有啥想法吗？

【英文标题】：SQL Statement Running Slow - Any Ideas?SQL 语句运行缓慢 - 有什么想法吗？

【发布时间】：2013-09-25 09:10:01

【问题描述】：

不确定下面的语句是怎么回事。运行时间超过 5 分钟 - 当我将主语句包装在一个 select 中并将其别名为 temp 时，它才开始显示出真正的减速，但这样做从未减慢我的任何其他 SQL。有任何想法吗？我在 mysql 上。

select *

from (
  select emp.company 
  , emp.employee
  , emp.last_name "Last Name"
  , emp.first_name "First Name"
  , emp.middle_init "Middle Initial"
  , trim(last_name) || ', ' || trim(first_name) ||  
      decode(trim(middle_init),'','',' '|| trim(middle_init)|| '.') "Full Name"
  , emp.emp_status
  , emp.work_country "Country"
  , (select
    (Case pgs.group_name
      when 'HREMEA' then 'EMEA'
      when 'HRNAMER' then 'NA'
      when 'HRLAMER' then 'LA'
      when 'HRAPAC' then'APAC'
      end)  
      from lawson.pgselect pgs 
      where pgs.begin_value = emp.work_country
        and pgs.company = 1 
        and pgs.group_name in ('HREMEA','HRNAMER','HRLAMER','HRAPAC') ) "Region"
  , emp.department "Department"
  , (select trim(r_name)  
      from lawson.deptcode dpt  
      where company = emp.company  
        and trim(process_level) = trim(emp.process_level) 
        and trim(department) = trim(emp.department)) "Department Description"
  , emp.job_code "Job Code"
  , (select description 
      from lawson.jobcode jbc 
        where company = emp.company 
        and job_code = emp.job_code) "Job Title"
  , emp.supervisor
  , (select trim(last_name) || ', ' || trim(first_name) ||  
      decode(trim(middle_init),'','',' '|| trim(middle_init)|| '.') 
        from lawson.employee supv 
        where supv.company = 1 
          and supv.position = emp.supervisor 
          and term_date = '01-JAN-1700') "Supervisor Name"
  , (select a_field 
      from lawson.hrempusf 
        where company = emp.company and employee = emp.employee 
        and field_key = '99') "Alt Mgr"
  , (select a_field 
      from lawson.hrempusf 
        where company = emp.company and employee = emp.employee 
        and field_key = '79') "TE Proxy Approver"
  , (select a_field 
      from lawson.hrempusf 
        where company = emp.company and employee = emp.employee 
        and field_key = '76') "Time Entry Proxy 1"
  , (select a_field 
      from lawson.hrempusf 
        where company = emp.company and employee = emp.employee
        and field_key = '77') "Time Entry Proxy 2"
  from lawson.employee emp 
  where term_date = '01-JAN-1700'
  ) temp

where temp."TE Proxy Approver" <> ' '
   or temp."Time Entry Proxy 1" <> ' '
   or temp."Time Entry Proxy 2" <> ' '

【问题讨论】：

我认为使用JOIN 而不是从同一个表中进行多次选择可以加快速度。取决于数据量。“慢”是相对的 - 如果查询每隔几个小时运行一次，5 分钟并不是很长的时间。

【参考方案1】：

用join替换子查询，并确保表有索引

【讨论】：

我会试一试的。谢谢。

【参考方案2】：

快速查看您的查询：

子查询太多，用连接替换它们 在 field_key 和 term_date 上缺少索引 不需要临时子查询，将 where 移到主查询中（联接中的数据更少）

【讨论】：

如果我将 where 移到我的主查询中，我无法指定“TE Proxy Approver”等，因为它们是别名......还是我遗漏了什么？

您应该进行内部联接而不是子查询，并在联接条件中指定条件