ORA-00600: kmgs_pre_process_request_6

一朋友动态修改oracle内存导致实例down掉。

记录于此

Tue Aug  5 16:07:51 2008
Errors in file /u01/app/oracle/admin/billdb/bdump/billdb2_mman_7496.trc:
ORA-00600: internal error code, arguments: [kmgs_pre_process_request_6], [4], [8], [6], [3], [0xC0000003398B4198], [], []
Tue Aug  5 16:07:53 2008
Errors in file /u01/app/oracle/admin/billdb/bdump/billdb2_mman_7496.trc:
ORA-00600: internal error code, arguments: [kmgs_pre_process_request_6], [4], [8], [6], [3], [0xC0000003398B4198], [], []
Tue Aug  5 16:07:53 2008
Trace dumping is performing id=[cdmp_20080805160753]
Tue Aug  5 16:07:53 2008
MMAN: terminating instance due to error 822
Tue Aug  5 16:07:55 2008
Shutting down instance (abort)
License high water mark = 47
Tue Aug  5 16:07:58 2008
Instance terminated by MMAN, pid = 7496
Tue Aug  5 16:08:00 2008
Instance terminated by USER, pid = 23950
Tue Aug  5 16:08:09 2008
Starting ORACLE instance (normal)
LICENSE_MAX_SESSION = 0
LICENSE_SESSIONS_WARNING = 0
Interface type 1 lan3 10.1.1.0 configured from OCR for use as a cluster interconnect
Interface type 1 lan0 192.168.0.0 configured from OCR for use as  a public interface
Picked latch-free SCN scheme 3
WARNING: db_recovery_file_dest is same as db_create_file_dest
Autotune of undo retention is turned on.
LICENSE_MAX_USERS = 0
SYS auditing is disabled
ksdpec: called for event 13740 prior to event group initialization
Starting up ORACLE RDBMS Version: 10.2.0.3.0.
System parameters with non-default values:
  processes                = 1000
  sessions                 = 1105
  __shared_pool_size       = 2566914048
  shared_pool_size         = 1375731712
  __large_pool_size        = 117440512
  large_pool_size          = 117440512
  __java_pool_size         = 16777216
  java_pool_size           = 16777216
  __streams_pool_size      = 134217728
  streams_pool_size        = 100663296
  spfile                   = +DG_DATA_01/billdb/spfilebilldb.ora
  nls_language             = AMERICAN
  sga_target               = 12884901888
  control_files            = +DG_DATA_01/billdb/controlfile/current.261.633805011, +DG_DATA_01/billdb/controlfile/current.260.633805013
  db_block_size            = 8192
  __db_cache_size          = 10032775168
  db_cache_size            = 6442450944
  compatible               = 10.2.0.3.0
  log_archive_dest_1       = LOCATION=+DG_DATA_01/billdb/
  log_archive_format       = %t_%s_%r.dbf
  db_files                 = 500
  db_file_multiblock_read_count= 32
  cluster_database         = TRUE
  cluster_database_instances= 2
  db_create_file_dest      = +DG_DATA_01
  db_recovery_file_dest    = +DG_DATA_01
  db_recovery_file_dest_size= 21474836480
  thread                   = 2
  instance_number          = 2
  undo_management          = AUTO
  undo_tablespace          = UNDOTBS2
  remote_login_passwordfile= EXCLUSIVE
  db_domain                = fs
  global_names             = TRUE
  dispatchers              = (PROTOCOL=TCP) (SERVICE=billdbXDB)
  local_listener           = (ADDRESS = (PROTOCOL = TCP)(HOST = crs_vip2) (PORT = 1521))
  remote_listener          = BILLDB1
  smtp_out_server          = 

  job_queue_processes      = 10
  background_dump_dest     = /u01/app/oracle/admin/billdb/bdump
  user_dump_dest           = /u01/app/oracle/admin/billdb/udump
  core_dump_dest           = /u01/app/oracle/admin/billdb/cdump
  audit_file_dest          = /u01/app/oracle/admin/billdb/adump
  db_name                  = billdb
  open_cursors             = 1000
  pga_aggregate_target     = 1702887424
Cluster communication is configured to use the following interface(s) for this instance
  10.1.1.2
Tue Aug  5 16:08:11 2008
cluster interconnect IPC version:Oracle UDP/IP (generic)
IPC Vendor 1 proto 2
PMON started with pid=2, OS id=24388
DIAG started with pid=3, OS id=24481
PSP0 started with pid=4, OS id=24513
LMON started with pid=5, OS id=24516
LMD0 started with pid=6, OS id=24518
LMS0 started with pid=7, OS id=24520
LMS1 started with pid=9, OS id=24526
MMAN started with pid=10, OS id=24555
DBW0 started with pid=11, OS id=24557
LGWR started with pid=12, OS id=24559
CKPT started with pid=13, OS id=24561
SMON started with pid=14, OS id=24563
RECO started with pid=15, OS id=24592
CJQ0 started with pid=16, OS id=24594
MMON started with pid=17, OS id=24596
Tue Aug  5 16:08:19 2008
starting up 1 dispatcher(s) for network address '(ADDRESS=(PARTIAL=YES)(PROTOCOL=TCP))'...
MMNL started with pid=18, OS id=24598
Tue Aug  5 16:08:19 2008
starting up 1 shared server(s) ...
Tue Aug  5 16:08:20 2008
lmon registered with NM - instance id 2 (internal mem no 1)
Tue Aug  5 16:08:21 2008
Reconfiguration started (old inc 0, new inc 8)
List of nodes:
 0 1
 Global Resource Directory frozen
* allocate domain 0, invalid = TRUE
 Communication channels reestablished
 * domain 0 valid = 0 according to instance 0
Tue Aug  5 16:08:21 2008
 Master broadcasted resource hash value bitmaps
 Non-local Process blocks cleaned out
Tue Aug  5 16:08:21 2008
 LMS 1: 0 GCS shadows cancelled, 0 closed
Tue Aug  5 16:08:21 2008
 LMS 0: 0 GCS shadows cancelled, 0 closed
 Set master node info
 Submitted all remote-enqueue requests
 Dwn-cvts replayed, VALBLKs dubious
 All grantable enqueues granted
Tue Aug  5 16:08:22 2008
 LMS 0: 0 GCS shadows traversed, 0 replayed
Tue Aug  5 16:08:22 2008
 LMS 1: 0 GCS shadows traversed, 0 replayed
Tue Aug  5 16:08:22 2008
 Submitted all GCS remote-cache requests
 Fix write in gcs resources
Reconfiguration complete
LCK0 started with pid=21, OS id=24780
Tue Aug  5 16:08:25 2008
ALTER DATABASE   MOUNT
Tue Aug  5 16:08:25 2008
Starting background process ASMB
ASMB started with pid=23, OS id=24810
Starting background process RBAL
RBAL started with pid=24, OS id=24814
Tue Aug  5 16:08:30 2008
SUCCESS: diskgroup DG_DATA_01 was mounted
Tue Aug  5 16:08:34 2008
Setting recovery target incarnation to 2
Tue Aug  5 16:08:34 2008
Successful mount of redo thread 2, with mount id 1587882380
Tue Aug  5 16:08:34 2008
Database mounted in Shared Mode (CLUSTER_DATABASE=TRUE)
Completed: ALTER DATABASE   MOUNT
Tue Aug  5 16:08:35 2008
ALTER DATABASE OPEN
Picked broadcast on commit scheme to generate SCNs
Tue Aug  5 16:08:35 2008
SUCCESS: diskgroup DG_DATA_03 was mounted
SUCCESS: diskgroup DG_DATA_02 was mounted
Tue Aug  5 16:08:36 2008
Starting ORACLE instance (normal)
Tue Aug  5 16:09:02 2008
LGWR: STARTING ARCH PROCESSES
ARC0 started with pid=8, OS id=25845
Tue Aug  5 16:09:02 2008
ARC0: Archival started
ARC1: Archival started
LGWR: STARTING ARCH PROCESSES COMPLETE
ARC1 started with pid=27, OS id=25847
Tue Aug  5 16:09:02 2008
Thread 2 opened at log sequence 38038
  Current log# 3 seq# 38038 mem# 0: +DG_DATA_01/billdb/onlinelog/group_3.268.633805093
Tue Aug  5 16:09:02 2008
ARC0: Becoming the 'no FAL' ARCH
ARC0: Becoming the 'no SRL' ARCH
Tue Aug  5 16:09:02 2008
ARC1: Becoming the heartbeat ARCH
Tue Aug  5 16:09:03 2008
Successful open of redo thread 2
Tue Aug  5 16:09:03 2008
MTTR advisory is disabled because FAST_START_MTTR_TARGET is not set
Tue Aug  5 16:09:03 2008
SMON: enabling cache recovery
Tue Aug  5 16:09:03 2008
Successfully onlined Undo Tablespace 5.
Tue Aug  5 16:09:03 2008
SMON: enabling tx recovery
Tue Aug  5 16:09:03 2008
Database Characterset is ZHS16CGB231280
replication_dependency_tracking turned off (no async multimaster replication found)
Tue Aug  5 16:09:05 2008
Streams APPLY A001 started with pid=29, OS id=25909
Tue Aug  5 16:09:08 2008
Starting background process QMNC
QMNC started with pid=30, OS id=25992
Tue Aug  5 16:09:11 2008
Completed: ALTER DATABASE OPEN

 

从metalink发现这是Oracle的另一个Bug 4433838

总之在rac环境下修改内存参数很容易引起实例down掉。

在动态修改内存的时候三思而后行，如果非要修改一定要在系统负载低的时候，否则你运气足够好的话，问题就出现了。并不是每一次都很幸运。

 

Subject:	Bug 4433838 - Resizing cache crashes instance with OERI[kmgs_pre_process_request_6]
	:	Note:4433838.8	Type:	PATCH
	Last Revision Date:	09-MAY-2008	Status:	PUBLISHED

 Click for details of sections in this note.

Bug 4433838  Resizing cache crashes instance with OERI[kmgs_pre_process_request_6]

 This note gives a brief overview of bug 4433838.

Affects:

Product (Component)	Oracle Server (Rdbms)
Range of versions believed to be affected	Versions < 11
Versions confirmed as being affected	10.2.0.3
Platforms affected	Generic (all / most platforms affected)

Fixed:

This issue is fixed in

10.2.0.4 (Server Patch Set) 11.1.0.6 (Base Release)

Symptoms:	Related To:
Internal Error May Occur (ORA-600) ORA-600 [kmgs_pre_process_request_6]	Automatic Memory Management

Description

Resizing the cache can crash the instance with 
ORA-600[kmgs_pre_process_request_6]. This can occur if SGA_TARGET
is set > 4Gb with the low order word of the size = 0.

Workaround: 
  Use a different sga_target such that lower order word is not zero.

---