oracle ~ $ oerr ORA 27102
27102, 00000, "out of memory"
// *Cause: Out of memory
// *Action: Consult the trace file for details

#below trace log file:
.......
ALTER SYSTEM SET sga_max_size='4096' SCOPE=SPFILE;
Mon Oct 19 20:20:57 2009
ALTER SYSTEM SET sga_max_size='4096M' SCOPE=SPFILE;
Mon Oct 19 20:21:37 2009
Shutting down instance (abort)
License high water mark = 3
Instance terminated by USER, pid = 26309
Mon Oct 19 20:21:39 2009
Starting ORACLE instance (normal)
Mon Oct 19 20:22:21 2009
Starting ORACLE instance (normal)
Mon Oct 19 20:23:05 2009
Starting ORACLE instance (normal)
Mon Oct 19 20:23:15 2009
Starting ORACLE instance (normal)
Mon Oct 19 20:24:14 2009
Starting ORACLE instance (normal)
LICENSE_MAX_SESSION = 0
LICENSE_SESSIONS_WARNING = 0
Picked latch-free SCN scheme 2
Using LOG_ARCHIVE_DEST_10 parameter default value as USE_DB_RECOVERY_FILE_DEST
Autotune of undo retention is turned on. 
IMODE=BR
ILAT =18
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                = 150
  __shared_pool_size       = 285212672
  __large_pool_size        = 16777216
  __java_pool_size         = 16777216
  __streams_pool_size      = 0
  nls_language             = AMERICAN
  sga_target               = 1191182336
  control_files            = /home/oracle/oradata/MREP/control01.ctl, /home/oracle/oradata/MREP/control02.ctl, /home/oracle/oradata/MREP/control03.ctl
  db_block_size            = 8192
  __db_cache_size          = 855638016
  compatible               = 10.2.0.1.0
  db_file_multiblock_read_count= 16
  db_recovery_file_dest    = /home/oracle/flash_recovery_area
  db_recovery_file_dest_size= 8489271296
  undo_management          = AUTO
  undo_tablespace          = UNDOTBS1
  remote_login_passwordfile= EXCLUSIVE
  db_domain                = 
  dispatchers              = (PROTOCOL=TCP) (SERVICE=MREPXDB)
  job_queue_processes      = 10
  background_dump_dest     = /home/oracle/admin/MREP/bdump
  user_dump_dest           = /home/oracle/admin/MREP/udump
  core_dump_dest           = /home/oracle/admin/MREP/cdump
  audit_file_dest          = /home/oracle/admin/MREP/adump
  db_name                  = MREP
  open_cursors             = 300
  pga_aggregate_target     = 392167424
PMON started with pid=2, OS id=26478
PSP0 started with pid=3, OS id=26480
MMAN started with pid=4, OS id=26482
DBW0 started with pid=5, OS id=26484
LGWR started with pid=6, OS id=26486
CKPT started with pid=7, OS id=26488
SMON started with pid=8, OS id=26490
RECO started with pid=9, OS id=26492
CJQ0 started with pid=10, OS id=26494
MMON started with pid=11, OS id=26496
Mon Oct 19 20:24:15 2009
starting up 1 dispatcher(s) for network address '(ADDRESS=(PARTIAL=YES)(PROTOCOL=TCP))'...
MMNL started with pid=12, OS id=26498
Mon Oct 19 20:24:15 2009
starting up 1 shared server(s) ...
Mon Oct 19 20:24:16 2009
ALTER DATABASE   MOUNT
Mon Oct 19 20:24:20 2009
Setting recovery target incarnation to 2
Mon Oct 19 20:24:20 2009
Successful mount of redo thread 1, with mount id 568876272
Mon Oct 19 20:24:20 2009
Database mounted in Exclusive Mode
Completed: ALTER DATABASE   MOUNT
Mon Oct 19 20:24:20 2009
ALTER DATABASE OPEN
Mon Oct 19 20:24:20 2009
Beginning crash recovery of 1 threads
 parallel recovery started with 7 processes
Mon Oct 19 20:24:20 2009
Started redo scan
Mon Oct 19 20:24:21 2009
Completed redo scan
 220 redo blocks read, 79 data blocks need recovery
Mon Oct 19 20:24:21 2009
Started redo application at
 Thread 1: logseq 8675, block 1220566
Mon Oct 19 20:24:21 2009
Recovery of Online Redo Log: Thread 1 Group 4 Seq 8675 Reading mem 0
  Mem# 0 errs 0: /disk1/oraclets/oraredo/log04a.rdo
  Mem# 1 errs 0: /disk1/oraclets/oraredo/log04b.rdo
Mon Oct 19 20:24:21 2009
Completed redo application
Mon Oct 19 20:24:21 2009
Completed crash recovery at
 Thread 1: logseq 8675, block 1220786, scn 10393969719
 79 data blocks read, 79 data blocks written, 220 redo blocks read
Mon Oct 19 20:24:21 2009
Thread 1 advanced to log sequence 8676
Thread 1 opened at log sequence 8676
  Current log# 3 seq# 8676 mem# 0: /home/oracle/oradata/MREP/redo03.log
Successful open of redo thread 1
Mon Oct 19 20:24:21 2009
MTTR advisory is disabled because FAST_START_MTTR_TARGET is not set
Mon Oct 19 20:24:21 2009
SMON: enabling cache recovery
Mon Oct 19 20:24:22 2009
Successfully onlined Undo Tablespace 1.
Mon Oct 19 20:24:22 2009
SMON: enabling tx recovery
Mon Oct 19 20:24:22 2009
Database Characterset is WE8ISO8859P1
replication_dependency_tracking turned off (no async multimaster replication found)
Starting background process QMNC
QMNC started with pid=23, OS id=26520
Mon Oct 19 20:24:27 2009
db_recovery_file_dest_size of 8096 MB is 0.00% used. This is a
user-specified limit on the amount of space that will be used by this
database for recovery-related files, and does not reflect the amount of
space available in the underlying filesystem or ASM diskgroup.
Mon Oct 19 20:24:27 2009
Completed: ALTER DATABASE OPEN
Mon Oct 19 22:24:29 2009
Thread 1 advanced to log sequence 8677
  Current log# 1 seq# 8677 mem# 0: /home/oracle/oradata/MREP/redo01.log