Question

我有以下查询，它给了我正确的结果，但代价超过3秒。处理结果非常慢。是否有任何优化技术可以应用于此查询以改善性能/响应时间，例如＆lt; 1秒。

select dept, location , count(*) total from 
(
    select empName, dept, location  from companyTable t1 where  (position = 'manager' or position = 'supervisor' )and  dateTime in 
    (
        select max(dateTime) from companyTable t2 where t2.empName = t1.empName and location = 'Dallas'   and dateTime like '2016:09:05%' 
    )
    union all

    select empName, dept, location from companyTable t10 where  (position = 'manager' or position = 'supervisor')and dateTime in 
    (
        select min(dateTime) from companyTable t2 where t2.empName = t10.empName and location <> 'Dallas'  and dateTime like '2016:09:05%' 
    )
) 
group by dept, empName having count(empName) > 1;

SCHEMA

CREATE TABLE companyTable
( dept varchar(32),location varchar(30),empName varchar(32), position varchar(30), dateTime time,
Primary key (dept, empName ,location ,position) );

EXPLAIN QUERY PLAN

2|0|0|SCAN TABLE companyTable AS t1
2|0|0|EXECUTE CORRELATED LIST SUBQUERY 3
3|0|0|SEARCH TABLE companyTable AS t2 USING INDEX sqlite_autoindex_companyTable_1 (dept=?)
4|0|0|SCAN TABLE companyTable AS t10
4|0|0|EXECUTE CORRELATED LIST SUBQUERY 5
5|0|0|SEARCH TABLE companyTable AS t2 USING INDEX sqlite_autoindex_companyTable_1 (dept=?)
1|0|0|COMPOUND SUBQUERIES 2 AND 4 (UNION ALL)
0|0|0|SCAN SUBQUERY 1
0|0|0|USE TEMP B-TREE FOR GROUP BY

sqlite3查询响应很慢

0 个答案: