ROLLUP会根据GROUP BY后面的字段从右到左逐步以去掉右边一个字段，逐步向上累计求和，
CUBE则会对GROUP BY后面的字段所有的组合进行分组求和。
CUBE eg1：
SELECT department_id, job_id, SUM (salary)
FROM employees
WHERE department_id < 60
GROUP BY CUBE (job_id, department_id);
DEPARTMENT_ID JOB_ID SUM(SALARY)
------------- ---------- -----------
211200
10 4400
20 19000
30 24900
40 6500
50 156400
HR_REP 6500
40 HR_REP 6500
MK_MAN 13000
20 MK_MAN 13000
MK_REP 6000
20 MK_REP 6000
PU_MAN 11000
30 PU_MAN 11000
ST_MAN 36400
50 ST_MAN 36400
AD_ASST 4400
10 AD_ASST 4400
PU_CLERK 13900
30 PU_CLERK 13900
SH_CLERK 64300
DEPARTMENT_ID JOB_ID SUM(SALARY)
------------- ---------- -----------
50 SH_CLERK 64300
ST_CLERK 55700
50 ST_CLERK 55700
24 rows selected.
CUBE eg2：
SELECT department_id, job_id, SUM (salary)
FROM employees
WHERE department_id < 60
GROUP BY CUBE (department_id, job_id);
DEPARTMENT_ID JOB_ID SUM(SALARY)
------------- ---------- -----------
211200
HR_REP 6500
MK_MAN 13000
MK_REP 6000
PU_MAN 11000
ST_MAN 36400
AD_ASST 4400
PU_CLERK 13900
SH_CLERK 64300
ST_CLERK 55700
10 4400
10 AD_ASST 4400
20 19000
20 MK_MAN 13000
20 MK_REP 6000
30 24900
30 PU_MAN 11000
30 PU_CLERK 13900
40 6500
40 HR_REP 6500
50 156400
DEPARTMENT_ID JOB_ID SUM(SALARY)
------------- ---------- -----------
50 ST_MAN 36400
50 SH_CLERK 64300
50 ST_CLERK 55700
24 rows selected.
本篇文章来源于 Linux公社网站() 原文链接：http:///Linux/2011-08/40467.htm
来自 “ ITPUB博客 ” ，链接：https://blog.itpub.net/16860121/viewspace-713679/ ，如需转载，请注明出处，否则将追究法律责任。
