ORA-00600 [keltnfy-ldmInit] internal error issue

在数据库运行的情况下因为硬件更换修改了IP地址，数据库停掉后就无法再次startup

alert.log

Sat Sep  1 06:21:02 2012
Starting ORACLE instance (normal)
LICENSE_MAX_SESSION = 0
LICENSE_SESSIONS_WARNING = 0
Picked latch-free SCN scheme 1
Using LOG_ARCHIVE_DEST_10 parameter default value as USE_DB_RECOVERY_FILE_DEST
Autotune of undo retention is turned on.
IMODE=BR
ILAT =242
LICENSE_MAX_USERS = 0
SYS auditing is disabled
ksdpec: called for event 13740 prior to event group initialization
Starting up ORACLE RDBMS Version: 10.2.0.1.0.
System parameters with non-default values:
  processes                = 2000
  sessions                 = 2205
  __shared_pool_size       = 889192448
  shared_pool_size         = 889192448
  __large_pool_size        = 167772160
  large_pool_size          = 167772160
  java_pool_size           = 218103808
  streams_pool_size        = 100663296
  control_files            = /oradata/NCME/controlfile/o1_mf_80dd16t9_.ctl, /oracle/flash_recovery_area/NCME/controlfile/o1_mf_80dd16vn_.ct
l
  db_block_size            = 8192
  __db_cache_size          = 6090129408
  db_cache_size            = 6090129408
  db_keep_cache_size       = 67108864
  compatible               = 10.2.0.1.0
  db_file_multiblock_read_count= 16
  db_create_file_dest      = /oradata
  db_recovery_file_dest    = /oracle/flash_recovery_area
  db_recovery_file_dest_size= 21474836480
  undo_management          = AUTO
  undo_tablespace          = UNDOTBS1
  remote_login_passwordfile= EXCLUSIVE
  db_domain                =
  dispatchers              = (PROTOCOL=TCP) (SERVICE=ncmeXDB)
  job_queue_processes      = 10
  background_dump_dest     = /oracle/admin/ncme/bdump
  user_dump_dest           = /oracle/admin/ncme/udump
  core_dump_dest           = /oracle/admin/ncme/cdump
  audit_file_dest          = /oracle/admin/ncme/adump
  db_name                  = ncme
  open_cursors             = 300
  optimizer_index_cost_adj = 70
  pga_aggregate_target     = 2097152000
Sat Sep  1 06:21:02 2012
Errors in file /oracle/admin/ncme/udump/ncme_ora_4494.trc:
ORA-00600: internal error code, arguments: [keltnfy-ldmInit], [46], [1], [], [], [], [], []
USER: terminating instance due to error 600
Instance terminated by USER, pid = 4494

trace file

Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - 64bit Production
With the Partitioning, OLAP and Data Mining options
ORACLE_HOME = /oracle/product/10.2.0
System name:    Linux
Node name:      crm-db
Release:        2.6.9-67.ELsmp
Version:        #1 SMP Fri Nov 16 12:49:06 EST 2007
Machine:        x86_64
Instance name: ncme
Redo thread mounted by this instance: 0Oracle process number: 0
Unix process pid: 4494, image: oracle@crm-db

KCCDEBUG_LEVEL = 0
Dynamic strand is set to TRUE
Running with 2 shared and 242 private strand(s). Zero-copy redo is FALSE
*** 2012-09-01 06:21:02.996
ksedmp: internal or fatal error
ORA-00600: internal error code, arguments: [keltnfy-ldmInit], [46], [1], [], [], [], [], []
Current SQL information unavailable - no session.
----- Call Stack Trace -----
calling              call     entry                argument values in hex
location             type     point                (? means dubious value)
-------------------- -------- -------------------- ----------------------------
ksedst()+31          call     ksedst1()            000000000 ? 000000001 ?
                                                   000000000 ? 000000000 ?
                                                   000000000 ? 000000001 ?
ksedmp()+610         call     ksedst()             000000000 ? 000000001 ?
                                                   000000000 ? 000000000 ?
                                                   000000000 ? 000000001 ?
ksfdmp()+21          call     ksedmp()             000000003 ? 000000001 ?
                                                   000000000 ? 000000000 ?
                                                   000000000 ? 000000001 ?
kgerinv()+161        call     ksfdmp()             000000003 ? 000000001 ?
                                                   000000000 ? 000000000 ?
                                                   000000000 ? 000000001 ?
kgeasnmierr()+163    call     kgerinv()            006469D40 ? 2A971DB118 ?
                                                   000000000 ? 000000000 ?
                                                   000000001 ? 000000001 ?
keltnfy()+286        call     kgeasnmierr()        006469D40 ? 2A971DB118 ?
                                                   000000000 ? 000000000 ?
                                                   000000000 ? 00000002E ?
kscnfy()+587         call     keltnfy()            006469D40 ? 7FBFFFB1D0 ?
                                                   000000000 ? 000000000 ?
                                                   000000000 ? 00000002E ?
ksucrp()+942         call     kscnfy()             00000000F ? 221D21218 ?
                                                   000000000 ? 000000000 ?
                                                   000000000 ? 00000002E ?
opistr_real()+1216   call     ksucrp()             000000000 ? 221D21218 ?
                                                   000000002 ? 000000000 ?

原因MOS也有说，就是当oracle因修改机器名或IP后无法确认机器的BUG，在10201 中存在，在10204中已fixed，当前的环境正好是centos5+oracle10201

解决办法
$hostname crm-db –hostname

vi /etc/hosts
127.0.0.1 localhost.localdomain localhost
192.168.212.100 crm-db — 新增加

$lsnrctl stop
$lsnrctl start

OK,再次打开数据库一切正常