oracle資料庫ORA-600 k2vcbk_2 故障恢複

 

有朋友找到我說他們資料庫無法啟動，資料庫啟動報ORA-600[k2vcbk_2]錯誤，資料庫版本為11.2.0.2 RAC，作業系統是AIX 6.1
SQL> recover database;
Media recovery complete.
SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-01092: ORACLE instance terminated. Disconnection forced
ORA-00600: internal error code, arguments: [k2vcbk_2], [], [], [], [], [], [],
[], [], [], [], []
Process ID: 7930020
Session ID: 49 Serial number: 14761
資料庫節點1日誌
Mon Sep 21 15:45:41 2015
Thread 1 advanced to log sequence 54076 (LGWR switch)
 Current log# 13 seq# 54076 mem# 0: +DG01/xifenfei/onlinelog/group_13.332.779459035
 Current log# 13 seq# 54076 mem# 1: +DG01/xifenfei/onlinelog/group_13.344.779582621
Mon Sep 21 15:45:44 2015
Archived Log entry 74655 added for thread 1 sequence 54075 ID 0x5a0bc0e1 dest 1:
Mon Sep 21 15:56:18 2015
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei1/trace/xifenfei1_ora_18088342.trc  (incident=184348):
ORA-00600: 內部錯誤碼, 參數: [kturPOTS_0], [], [], [], [], [], [], [], [], [], [], []
Incident details in: /oracle/diag/rdbms/xifenfei/xifenfei1/incident/incdir_184348/xifenfei1_ora_18088342_i184348.trc
Mon Sep 21 15:56:34 2015
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Error 600 trapped in 2PC on transaction 7.16.120119. Cleaning up.
Error stack returned to user:
ORA-00600: 內部錯誤碼, 參數: [kturPOTS_0], [], [], [], [], [], [], [], [], [], [], []
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei1/trace/xifenfei1_ora_18088342.trc  (incident=184349):
ORA-00603: ORACLE 伺服器會話因致命錯誤而終止
ORA-00600: 內部錯誤碼, 參數: [kturPOTS_0], [], [], [], [], [], [], [], [], [], [], []
Mon Sep 21 15:56:34 2015
Dumping diagnostic data in directory=[cdmp_20150921155634], requested by (instance=1, osid=18088342), summary=[incident=184348].
Incident details in: /oracle/diag/rdbms/xifenfei/xifenfei1/incident/incdir_184349/xifenfei1_ora_18088342_i184349.trc
Mon Sep 21 15:56:35 2015
Sweep [inc][184349]: completed
Sweep [inc][184348]: completed
Sweep [inc2][184348]: completed
opiodr aborting process unknown ospid (18088342) as a result of ORA-603
Mon Sep 21 15:57:12 2015
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei1/trace/xifenfei1_smon_7536810.trc  (incident=184274):
ORA-00600: internal error code, arguments: [k2vcbk_2], [], [], [], [], [], [], [], [], [], [], []
Incident details in: /oracle/diag/rdbms/xifenfei/xifenfei1/incident/incdir_184274/xifenfei1_smon_7536810_i184274.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Mon Sep 21 15:57:16 2015
Dumping diagnostic data in directory=[cdmp_20150921155716], requested by (instance=1, osid=7536810 (SMON)), summary=[incident=184274].
Fatal internal error happened while SMON was doing active transaction recovery.
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei1/trace/xifenfei1_smon_7536810.trc:
ORA-00600: internal error code, arguments: [k2vcbk_2], [], [], [], [], [], [], [], [], [], [], []
SMON (ospid: 7536810): terminating the instance due to error 474
Mon Sep 21 15:57:18 2015
ORA-1092 : opitsk aborting process
資料庫節點2日誌
Mon Sep 21 15:21:50 2015
Archived Log entry 74653 added for thread 2 sequence 23559 ID 0x5a0bc0e1 dest 1:
Mon Sep 21 15:44:28 2015
Thread 2 advanced to log sequence 23561 (LGWR switch)
 Current log# 12 seq# 23561 mem# 0: +DG01/xifenfei/onlinelog/group_12.338.779457003
 Current log# 12 seq# 23561 mem# 1: +DG01/xifenfei/onlinelog/group_12.265.779582493
Mon Sep 21 15:44:31 2015
Archived Log entry 74654 added for thread 2 sequence 23560 ID 0x5a0bc0e1 dest 1:
Mon Sep 21 15:45:31 2015
DISTRIB TRAN xifenfei.1ebab0a5.20.3.1533822
 is local tran 20.3.1533822 (hex=14.03.17677e)
 insert pending committed tran, scn=14590688068086 (hex=d45.28c781f6)
Mon Sep 21 15:45:31 2015
DISTRIB TRAN xifenfei.1ebab0a5.20.3.1533822
 is local tran 20.3.1533822 (hex=14.03.17677e))
 delete pending committed tran, scn=14590688068086 (hex=d45.28c781f6)
Mon Sep 21 15:56:35 2015
Dumping diagnostic data in directory=[cdmp_20150921155634], requested by (instance=1, osid=18088342), summary=[incident=184348].
Mon Sep 21 15:57:10 2015
Error 3135 trapped in 2PC on transaction 20.11.1534704. Cleaning up.
Error stack returned to user:
ORA-03135: 串連失去聯絡
opidcl aborting process unknown ospid (9175532) as a result of ORA-604
Mon Sep 21 15:57:17 2015
Dumping diagnostic data in directory=[cdmp_20150921155716], requested by (instance=1, osid=7536810 (SMON)), summary=[incident=184274].
Mon Sep 21 15:57:23 2015
Reconfiguration started (old inc 18, new inc 20)
List of instances:
2 (myinst: 2)
Global Resource Directory frozen
* dead instance detected - domain 0 invalid = TRUE
Communication channels reestablished
Master broadcasted resource hash value bitmaps
Non-local Process blocks cleaned out
Mon Sep 21 15:57:23 2015
LMS 2: 3 GCS shadows cancelled, 1 closed, 0 Xw survived
Mon Sep 21 15:57:23 2015
LMS 0: 2 GCS shadows cancelled, 0 closed, 0 Xw survived
Mon Sep 21 15:57:23 2015
LMS 1: 3 GCS shadows cancelled, 1 closed, 0 Xw survived
Set master node info
Submitted all remote-enqueue requests
Dwn-cvts replayed, VALBLKs dubious
All grantable enqueues granted
Post SMON to start 1st pass IR
Mon Sep 21 15:57:23 2015
minact-scn: Inst 2 is now the master inc#:20 mmon proc-id:6816208 status:0x7
minact-scn status: grec-scn:0x0000.00000000 gmin-scn:0x0d45.28c2bb5c gcalc-scn:0x0d45.28c3bd2e
minact-scn: master found reconf/inst-rec before recscn scan old-inc#:20 new-inc#:20
Mon Sep 21 15:57:23 2015
Instance recovery: looking for dead threads
Submitted all GCS remote-cache requests
Post SMON to start 1st pass IR
Fix write in gcs resources
Reconfiguration complete
Beginning instance recovery of 1 threads
parallel recovery started with 31 processes
Started redo scan
Completed redo scan
read 12626 KB redo, 1724 data blocks need recovery
Started redo application at
Thread 1: logseq 54076, block 184416
Recovery of Online Redo Log: Thread 1 Group 13 Seq 54076 Reading mem 0
 Mem# 0: +DG01/xifenfei/onlinelog/group_13.332.779459035
 Mem# 1: +DG01/xifenfei/onlinelog/group_13.344.779582621
Completed redo application of 9.78MB
Completed instance recovery at
Thread 1: logseq 54076, block 209669, scn 14590688357285
1633 data blocks read, 1794 data blocks written, 12626 redo k-bytes read
Thread 1 advanced to log sequence 54077 (thread recovery)
Mon Sep 21 15:57:33 2015
Error 3113 trapped in 2PC on transaction 21.18.1965522. Cleaning up.
Redo thread 1 internally disabled at seq 54077 (SMON)
Error stack returned to user:
ORA-02050: 交易處理 21.18.1965522 已回退, 某些遠端資料庫可能有問題
ORA-03113: 通訊通道的檔案結尾
ORA-02063: 緊接著 line (起自 ZSK)
Mon Sep 21 15:57:34 2015
Archived Log entry 74656 added for thread 1 sequence 54076 ID 0x5a0bc0e1 dest 1:
Mon Sep 21 15:57:34 2015
ARC0: Archiving disabled thread 1 sequence 54077
Archived Log entry 74657 added for thread 1 sequence 54077 ID 0x5a0bc0e1 dest 1:
Mon Sep 21 15:57:35 2015
Thread 2 advanced to log sequence 23562 (LGWR switch)
 Current log# 8 seq# 23562 mem# 0: +DG01/xifenfei/onlinelog/group_8.334.779456945
 Current log# 8 seq# 23562 mem# 1: +DG01/xifenfei/onlinelog/group_8.267.779582453
Mon Sep 21 15:57:36 2015
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei2/trace/xifenfei2_smon_6750672.trc  (incident=200218):
ORA-00600: internal error code, arguments: [k2vcbk_2], [], [], [], [], [], [], [], [], [], [], []
Incident details in: /oracle/diag/rdbms/xifenfei/xifenfei2/incident/incdir_200218/xifenfei2_smon_6750672_i200218.trc
Archived Log entry 74658 added for thread 2 sequence 23561 ID 0x5a0bc0e1 dest 1:
Mon Sep 21 15:57:38 2015
minact-scn: master continuing after IR
Mon Sep 21 15:57:41 2015
Dumping diagnostic data in directory=[cdmp_20150921155741], requested by (instance=2, osid=6750672 (SMON)), summary=[incident=200218].
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Fatal internal error happened while SMON was doing instance transaction recovery.
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei2/trace/xifenfei2_smon_6750672.trc:
ORA-00600: internal error code, arguments: [k2vcbk_2], [], [], [], [], [], [], [], [], [], [], []
SMON (ospid: 6750672): terminating the instance due to error 474
Mon Sep 21 15:57:41 2015
ORA-1092 : opitsk aborting process
Mon Sep 21 15:57:42 2015
ORA-1092 : opitsk aborting process
Mon Sep 21 15:57:42 2015
License high water mark = 291
Instance terminated by SMON, pid = 6750672
USER (ospid: 18874814): terminating the instance
通過資料庫日誌大概可以看出來，由於節點2的分布式事事務異常，而在11.2.0.2中，分散式交易跨節點，引起節點2的pmon清理異常事務，但是由於bug，使得異常事務無法被清理掉，從而引起節點1 crash,節點1 crash之後節點2進行恢複，也因為分散式交易bug，導致smon復原失敗，執行個體也crash。無法進行復原導致資料庫無法正常啟動，通過查詢mos發現定位到是Bug 10222544 ORA-600 [k2vpci_2] from multi-branch distributed transaction

對於這類問題，由於分布事務無法清理,處理方法就是找出來事務人工提交或者直接屏蔽掉這個事務解決該問題