Troubleshooting ORA-00600: [kpncxcc-1] from dblink & ORA-01801

前几日一数据库实例alert日志中出现了ORA-600 [kpncxcc-1], kpn开头的函数通常用于服务器之间通信, 从trace中也可以发现是一条调用了dblink的select, 在MOS未发现已知BUG.
ERROR:
ORA-00600: internal error code, arguments: [kpncxcc-1], [12], [5], [7], [], [], [], [], [], [], [], []

, , ,

Troubleshooting ORA-00600: [krccaub_2], [9], [10] & instance crash in Oracle 10.2.0.4 RAC

ORA-00600: internal error code, arguments: [krccaub_2], [9], [10], [], [], [], [], []
Sat Feb 25 05:14:31 2017
CTWR: terminating instance due to error 487

, ,

sqlplus login hang, syscal高, 频繁的实例重启 11.2.0.4 RAC on SunOS 5.10 (Solaris)

这是一套11.2.04 2节点4实例2数据库的CLUSTER, 每个节点上有2个实例, 每过一段时间其中1个节点上1个实例就会挂起hang死和实例驱逐重启, 主机环境是基于X86 64的Solaris OS (sunos 5.10)操作系统, 496 CPUs, 1TB的内存, 使用RAW裸设备存储…

, ,

Query V$SEGSTAT or V$SEGMENT_STATISTICS views slowly (慢)

昨天有个朋友在做性能监控时遇到了一个性能问题咨询我是否遇到过? 在他所有的数据库实例中只有一个节点查询V$SEGMENT_STATISTICS时几个小时都未出结果, 其实这个问题影响11.2.0.3 ,11.2.0.4多个版本及平台

, ,

当C语言的程序处理 chr(0) or ‘\0’ 时的ORA-01008 Case

前几日遇到一套用C语言封装的数据复制程序故障, 同步的数据变化数据存储在数据库,但同步的流程控制是封装好的, 今天突然日志中出现了ORA-01008: not all variables bound, 该错误是语法级错误当SQL中的实际付值数量小于绑定变量数时提示.

,

ora-01499 & ora-08103 caused by block corrupted or write loss a case

ORA-1499 is produced by statement “ANALIZE TABLE|CLUSTER VALIDATE STRUCTURE CASCADE” to report an inconsistency between a table or a cluster and its index where an index,ORA-8103 is caused by an invalid block type. The block header has an invalid block type or the block type inside the block is not expected

,

11.2 impdp ORA-31693 ORA-29913 ORA-27163 when the table has xmltype column

ORA-31693: Failed to load / unload the data object table “ANBOB” “SDTEMPETL” object is ignored because of the error.:
ORA-29913: error in executing the call ODCIEXTTABLEFETCH
ORA-27163: insufficient memory(内存不足)

,

Troubleshooting: ORA-00600: [kkpo_rcinfo_defstg:delseg], [xxxx] & ora-600 [25027] & ora-600 [ktadrprc-1]

因为某些原因数据字典表不一致,导到该表在查询或导出时都会提示ora-600 [kkpo_rcinfo_defstg:delseg] 错误,因为数据库使用延迟段创建,手动分配segment时提示ORA-600 [25027],对分区做MOVE时会提示ORA-600 [ktadrprc-1], 使用hcheck脚本检查会提示Orphaned TABPART$

,

Using ‘opatch lsinventory’ show patched is real? (看到的补丁信息真的靠谱么?)

去年在排查SCN 天花板问题修改相关的几个参数时,发现了这个问题,尤其如果是从别人手中接手的数据库,通常从opatch lsinventory 检查数据库的版本和补丁信息,在12C 的版本中也可以使用DBMS_QOPATCH, 但是有一个库发现之前别人挖了个坑实际安装补丁时应试是出错了,但是从opatch lsinventory查看两节点并没有区别 .。。。

,

ora-600 [ktbdchk1: bad dscn] and ora-8103 corrupted block

前两天一同事的数据库alert log不停的在刷出ora-600 [ktbdchk1: bad dscn]错误,影响的是一张table上的insert语句, 前不久该数据库存储出现过故障,从已知BUG中没有找到相似案例,这里只是简单的记录一下问题的处理过程。

,

Troubleshooting ORA-07445 [intel_fast_memcpy.A()+18]

前不久有个朋友从遇到了个问题,在数据库启动过程种出现ora-7445, 然后数据库CRASH, 环境11.1.0.7 windows 版本, 临时禁用了smon tx recovery打开数据库, 没有再继续跟踪,这里记录一下.
ORA-07445: exception encountered: core dump [intel_fast_memcpy.A()+18] [ACCESS_VIOLATION] [ADDR:0x19DC0000] [PC:0x52B428E] [UNABLE_TO_WRITE] []

,

ORA-00600 [ddfnetCFull-4], [Invalid Handle] when using shared dblink oracle 11.2.0.3

ORA-00600: internal error code, arguments: [ddfnetCFull-4], [Invalid Handle], [], [], [], [], [], [], [], [], [], []
—– Current SQL Statement for this session (sql_id=0wjqfgz9dqsjp) —–
select null from srp.interface_comm@srp1
where rowid = :plsqldev_rowid
for update nowait

Frequent generate a lot of cdmp* directories contain *bucket trace in bdump(二 ) root cause

之前记录过Frequent generate a lot of cdmp* directories contain *bucket trace in bdump , 当时根据现象与该bug 很相似,该bug解决的只是不生成cdump还是会生成trace 文件,所以要分析生成trace的根本原因。之前数据库在诊断问题时也启用过相关trace event(见下面) , 但是instance memroy 级别实例重启后不再生效

ORA-32701 error in alert.log M000 hang event ‘not in wait’ during flush AWR

环境是一套11.2.0.3 2nodes RAC on hpux-ia31, alert中出现ora-32701 hangmgr错误, 从trace文件中发现是m000进程是mmon的辅助进程,用于flush AWR相关数据,有一个wait event: enq: WF – contention, 这也是flush AWR数据时相关的enqueue等待,但是blocker进程是not in wait

Troubleshooting ORA-600 [kghstack_alloc] & ORA-600 [kponPurgeUnreachLoc-3]

ORA-00600: internal error code, arguments: [kghstack_alloc], [define handles], [], [], [], [], [], [], [], [], [], []
***** Current SQL Statement for this session (sql_id=5udxh0ykgshkb) *****
select location_name, max(r.reg_id) from reg$ r left outer join gv$subscr_registration_stats

,

ORA-01089 select fail over database link on RAC ADG Standby

前段时间配的一套11203 RAC ADG on EXADATA Machine的环境,在ADG 的standby side的node2 通过dblink查询时提示ora-1089错误,但是在node1 查询正常,DG的recover 进程是在node1 上..

,

Frequent generate a lot of cdmp* directories contain *bucket trace in bdump

最近有套库突然目录使用告警,发现TRACE 目录使用较高,发现在bdump 目录中每分钟都生成一个cdmp*目录,且每个目录中含大量的文件,这是一套oracle 11.2.0.3 rac on hpux 11.31的数据库,

Run root102.sh Node reboot when install10205 CRS PATCH & OUI-10021 & Oracle CRS stack is already configured

10.2.0.4 2nodes Oracle RAC ON HPUX 平台的环境, 计划升级到10.2.0.5, 前期升级clusterware 软件非常顺序,可是在跑root102.sh时还没跑完主机突然reboot, 尝试启crs 报错:The clusterware daemons are running from . But you are patching /oracle/product/10.2.0/crs10g

ORA-20011 ORA-12801 ORA-01722 when gather table statistics

在做一个普通的分区表(HEAP TABLE)收集统计信息时进程报错意外终止,后台日志未出现ora-600等内部错误,这是一套11.2.0.3.7 oracle RAC 2-nodes on HPUX-ia 11.31的运行环境,这里简单的记录一下排查过程。

SQL> exec DBMS_STATS.GATHER_TABLE_STATS (ownname => ‘ANBOB’ , tabname => ‘TAB_ERROR’ , cascade => true, estimate_percent => dbms_stats.auto_sample_size,method_opt=>’FOR TABLE FOR ALL COLUMNS SIZE REPEAT’, degree => 8,no_invalidate=>false);

*
ERROR at line 1:
ORA-20011: Approximate NDV failed: ORA-12801: error signaled in parallel query server P001, instance qdim1:im1 (1)
ORA-01722: invalid number

ora-7445 [LpxFSMSaxSE()+449] [SIGSEGV] when query with xmltype

ORA-07445: exception encountered: core dump [LpxFSMSaxSE()+449] [SIGSEGV] [ADDR:0xFD26CFB5] [PC:0x400000001117F881] [Address not mapped to object] []
—– Current SQL Statement for this session (sql_id=6qt1f1pg1dkzs) —–
SELECT UPPER(XMLType(CHR(60) || CHR(58) || CHR(113) || CHR(120) ||

,