Category Archives: oracle bugs

11g Bug 10082277 – Excessive allocation in PCUR or KGLH0 heap of “kkscsAddChildNo”

在11.2.0.2的数据库上,我们遇到了oracle 11g的Bug 10082277 Excessive allocation in PCUR or KGLH0 heap of “kkscsAddChildNo” (ORA-4031) Bug 10082277 Excessive allocation in PCUR or KGLH0 heap of “kkscsAddChildNo” (ORA-4031) This note gives a brief overview of bug 10082277. The content was last updated on: … Continue reading

Posted in oracle 11g, oracle bugs | 494 Comments

10203下log file parallel write统计信息不准确的问题

log file parallel write是很重要的一个等待事件,当lgwr将redo buffer中的内容写入到disk上去的时候会出现该等待事件。 注意不要和log file sequential read混淆了,log file sequential read是更新logfile header block,出现在日志切换或者归档的时候,等待进程可能是lgwr或者arch。 我发现在数据库从10203升级到10204后,该等待事件的等待次数变多了很多: 这样大幅度的提高蛮奇怪的,我在oracle-l上提出了该问题 http://www.freelists.org/post/oracle-l/log-file-parallel-write-wait-times-get-increased-a-lot-after-upgrading-from-10203-to-10204 Tanel Poder提到这是一个instrumentation bug,10203下面log file parallel write的统计信息不准确。而10204修复了这个问题,所以并不是10204有更多的等待事件,而是10204的统计更加准确了。 It’s more likely that you hit an instrumentation bug which didn’t record any log file parallel write … Continue reading

Posted in oracle bugs | Tagged , | 143 Comments

10203 connect_by性能问题

用户反映一条SQL语句在绑定某个值的时候性能急剧下降。一般只需要要几秒钟的执行时间,而当绑定这个值的时候需要好几分钟。 SQL使用了connect_by,添加gather_plan_statistics hint实际执行后,查看执行计划 select * from table(dbms_xplan.display_cursor(null,null,’ALLSTATS LAST’)); 为了显示方便,我省略了一些信息 ———————————————————————————————- | Id | Operation | Starts | A-Time | Buffers | Reads | Writes | ———————————————————————————————- | 1 | NESTED LOOPS OUTER | 1 |00:03:56.87 | 9654K| 136K| 67609 | … Continue reading

Posted in Life, oracle bugs | 79 Comments