SQL> select round(cpu.value / elapsed.value * 100, 2) "Parse CPU to Parse Elapsd %:"
2 from v$sysstat cpu, v$sysstat elapsed
3 where cpu.name = 'parse time cpu'
4 and elapsed.name = 'parse time elapsed';
Parse CPU to Parse Elapsd %:
----------------------------
32955.43
SQL> select * from v$sysstat where name like '%parse time%';
STATISTIC# NAME CLASS VALUE STAT_ID
---------- ------------------------------ ---------- ---------- ----------
336 parse time cpu 64 4299741367 206905303
337 parse time elapsed 64 13047396 1431595225
| 64 | Total CPU time used for parsing (hard and soft) in 10s of milliseconds | 3 | |
| 64 | Total elapsed time for parsing, in 10s of milliseconds. Subtract "parse time cpu" from the this statistic to determine the total waiting time for parse resources |
开了个SR 问Metalink, Metalink 工程师说 这个是计算误差导致的,我要他们详细的官方文档,没得, TNND 忽悠哥哥
本文通过SQL查询展示了Oracle数据库中解析CPU时间和解析耗时的关系,并针对一个高出预期的百分比进行了讨论。通过查询v$sysstat视图获取了parsetimecpu和parsetimeelapsed的具体数值,发现两者之间的计算结果出现了较大偏差。Metalink工程师认为这是由计算误差引起的。
6988

被折叠的 条评论
为什么被折叠?



