Sessions remains and increase cumulatively never close if you use DEDICATED with XA transaction + shared dblink.

A web application using Tuxedo always keep 50 long connection to local database which then issues a SELECT from local db to remote db over shared db link(it’s created last weekend). But to run for a period of time after the found the remote db sessions become full(ora-18)


Troubleshooting ORA 600 [ktspfmdb:objdchk_kcbnew_3] 案例

前段时间数据库出现了几次ORA 600 [ktspfmdb:objdchk_kcbnew_3]错误,引起该错误的是一条insert sql.

Troubleshooting these are too many scheduler jobs(owner sys) name like KWQICPOSTMSGDEL_nn in DB 11g

I faced a very interesting question today, An oracle database RAC database ON hpux one had to many scheduler jobs , And the job’s owner is sys, All jobs name all like ‘KWQICPOSTMSGDEL_’, All these jobs has no start date as well no interval.

How to estimate the cost if partition table missing partition level statistics ?(缺失分区表级统计信息如何计算多分区扫描的COST)

首先查看表的统计信息,分区表的统计信息分global stats(DBA_TABLE)和partition stats(DBA_TAB_PARTITION),发现该分区表是global stats 有,partition stats全为空, 查看该SQL发现有两个子游标有不同的执行计划

Troubleshooting ORA-600 [k2srec: should be another instance]

Problem Key: ORA 600 [k2srec: should be another instance]
Error: ORA-600 [k2srec: should be another instance] [2] [] [] [] [] [] [] [] [] [] [][00]
with the following stack :k2srec <- k2serv <- opiodr <- ttcpip

The Oracle Listener ofen restart and check fails on HPUX (监听频繁重启)

上个月有套库的listener总是自动stop, 因为是crs的资源又被agent重新拉起,前几次的监听故障中都提到过listener依赖于network资源, 以前也处理过网络打嗝可能会导致监听频繁重启的案例,本次参照该案例修改弱依赖的方式无法避免该问题