Oracle启动报错ORA-03113解决

环境：RHEL6.4 + Oracle 11.2.0.4

步骤摘要：
1.启动报错ORA-03113
2.查看alert日志查找原因
3.根据实际情况采取合理的措施，这里我们先增加闪回区大小，把库启动起来
4.制定对应的归档日志删除策略

1.启动报错ORA-03113

$ sqlplus / as sysdba

SQL*Plus: Release 11.2.0.4.0 Production on 星期六 2月 28 13:56:44 2015

Copyright (c) 1982, 2013, Oracle.  All rights reserved.

Connected to an idle instance.

SQL> startup mount
ORACLE instance started.

Total System Global Area 3206836224 bytes
Fixed Size                  2257520 bytes
Variable Size            2365590928 bytes
Database Buffers          822083584 bytes
Redo Buffers              16904192 bytes
Database mounted.
SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-03113: 通信通道的文件结尾
进程 ID: 13501
会话 ID: 853 序列号: 5

SQL>

2.查看alert日志查找原因

Sat Feb 28 13:56:52 2015
Starting ORACLE instance (normal)
LICENSE_MAX_SESSION = 0
LICENSE_SESSIONS_WARNING = 0
Initial number of CPU is 8
CELL communication is configured to use 0 interface(s):
CELL IP affinity details:
    NUMA status: non-NUMA system
    cellaffinity.ora status: N/A
CELL communication will use 1 IP group(s):
    Grp 0:
Picked latch-free SCN scheme 3
Using LOG_ARCHIVE_DEST_1 parameter default value as USE_DB_RECOVERY_FILE_DEST
Autotune of undo retention is turned on.
IMODE=BR
ILAT =249
LICENSE_MAX_USERS = 0
SYS auditing is disabled
Starting up:
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options.
ORACLE_HOME = /opt/app/oracle/product/11.2.0.4/dbhome_1
System name:    Linux
Node name:      JJFAB01
Release:        2.6.32-358.el6.x86_64
Version:        #1 SMP Tue Jan 29 11:47:41 EST 2013
Machine:        x86_64
VM name:        Xen Version: 4.1 (PVM)
Using parameter settings in server-side spfile /opt/app/oracle/product/11.2.0.4/dbhome_1/dbs/spfileJJFAB.ora
System parameters with non-default values:
  processes                = 1500
  sessions                = 2272
  memory_target            = 3G
  control_files            = "/opt/oradata/gxwj/control01.ctl"
  control_files            = "/opt/app/oracle/fast_recovery_area/gxwj/control02.ctl"
  db_block_size            = 8192
  compatible              = "11.2.0.0.0"
  db_recovery_file_dest    = "/opt/app/oracle/fast_recovery_area"
  db_recovery_file_dest_size= 4122M
  undo_tablespace          = "UNDOTBS1"
  remote_login_passwordfile= "EXCLUSIVE"
  db_domain                = ""
  dispatchers              = "(PROTOCOL=TCP) (SERVICE=gxwjXDB)"
  audit_file_dest          = "/opt/app/oracle/admin/gxwj/adump"
  audit_trail              = "DB"
  db_name                  = "JJFAB"
  open_cursors            = 300
  diagnostic_dest          = "/opt/app/oracle"
Sat Feb 28 13:56:54 2015
PMON started with pid=2, OS id=13461
Sat Feb 28 13:56:54 2015
PSP0 started with pid=3, OS id=13463
Sat Feb 28 13:56:55 2015
VKTM started with pid=4, OS id=13465 at elevated priority
VKTM running at (1)millisec precision with DBRM quantum (100)ms
Sat Feb 28 13:56:55 2015
GEN0 started with pid=5, OS id=13469
Sat Feb 28 13:56:55 2015
DIAG started with pid=6, OS id=13471
Sat Feb 28 13:56:55 2015
DBRM started with pid=7, OS id=13473
Sat Feb 28 13:56:55 2015
DIA0 started with pid=8, OS id=13475
Sat Feb 28 13:56:55 2015
MMAN started with pid=9, OS id=13477
Sat Feb 28 13:56:55 2015
DBW0 started with pid=10, OS id=13479
Sat Feb 28 13:56:55 2015
LGWR started with pid=11, OS id=13481
Sat Feb 28 13:56:55 2015
CKPT started with pid=12, OS id=13483
Sat Feb 28 13:56:55 2015
SMON started with pid=13, OS id=13485
Sat Feb 28 13:56:55 2015
RECO started with pid=14, OS id=13487
Sat Feb 28 13:56:55 2015
MMON started with pid=15, OS id=13489
starting up 1 dispatcher(s) for network address '(ADDRESS=(PARTIAL=YES)(PROTOCOL=TCP))'...
Sat Feb 28 13:56:55 2015
MMNL started with pid=16, OS id=13491
starting up 1 shared server(s) ...
ORACLE_BASE from environment = /opt/app/oracle
Sat Feb 28 13:56:56 2015
ALTER DATABASE  MOUNT
Successful mount of redo thread 1, with mount id 3498004520
Database mounted in Exclusive Mode
Lost write protection disabled
Completed: ALTER DATABASE  MOUNT
Sat Feb 28 13:57:48 2015
alter database open
Sat Feb 28 13:57:48 2015
LGWR: STARTING ARCH PROCESSES
Sat Feb 28 13:57:48 2015
ARC0 started with pid=20, OS id=13553
ARC0: Archival started
LGWR: STARTING ARCH PROCESSES COMPLETE
ARC0: STARTING ARCH PROCESSES
Sat Feb 28 13:57:49 2015
ARC1 started with pid=21, OS id=13555
Sat Feb 28 13:57:49 2015
ARC2 started with pid=22, OS id=13557
Errors in file /opt/app/oracle/diag/rdbms/jjfab/JJFAB/trace/JJFAB_ora_13501.trc:
ORA-19815: 警告: db_recovery_file_dest_size 字节 (共 4322230272 字节) 已使用 100.00%, 尚有 0 字节可用。
************************************************************************
You have following choices to free up space from recovery area:
1. Consider changing RMAN RETENTION POLICY. If you are using Data Guard,
  then consider changing RMAN ARCHIVELOG DELETION POLICY.
2. Back up files to tertiary device such as tape using RMAN
  BACKUP RECOVERY AREA command.
3. Add disk space and increase db_recovery_file_dest_size parameter to
  reflect the new space.
4. Delete unnecessary files using RMAN DELETE command. If an operating
  system command was used to delete files, then use RMAN CROSSCHECK and
  DELETE EXPIRED commands.
************************************************************************
ARCH: Error 19809 Creating archive log file to '/opt/app/oracle/fast_recovery_area/JJFAB/archivelog/2015_02_28/o1_mf_1_130_%u_.arc'
Errors in file /opt/app/oracle/diag/rdbms/jjfab/JJFAB/trace/JJFAB_ora_13501.trc:
ORA-16038: 日志 1 sequence# 130 无法归档
ORA-19809: 超出了恢复文件数的限制
ORA-00312: 联机日志 1 线程 1: '/opt/oradata/gxwj/redo01.log'
Sat Feb 28 13:57:49 2015
ARC3 started with pid=23, OS id=13559
USER (ospid: 13501): terminating the instance due to error 16038
Sat Feb 28 13:57:50 2015
System state dump requested by (instance=1, osid=13501), summary=[abnormal instance termination].
System State dumped to trace file /opt/app/oracle/diag/rdbms/jjfab/JJFAB/trace/JJFAB_diag_13471_20150228135750.trc
Dumping diagnostic data in directory=[cdmp_20150228135750], requested by (instance=1, osid=13501), summary=[abnormal instance termination].
Instance terminated by USER, pid = 13501

发现告警日志中已经明确给出了原因和建议。

3.根据实际情况采取合理的措施，这里我们先增加闪回区大小，把库启动起来。

SQL> startup mount
ORA-24324: 未初始化服务句柄
ORA-01041: 内部错误, hostdef 扩展名不存在
SQL> exit
Disconnected from Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
$ sqlplus / as sysdba

SQL*Plus: Release 11.2.0.4.0 Production on 星期六 2月 28 14:05:59 2015

Copyright (c) 1982, 2013, Oracle.  All rights reserved.

Connected to an idle instance.

SQL> startup mount
ORACLE instance started.

Total System Global Area 3206836224 bytes
Fixed Size                  2257520 bytes
Variable Size            2365590928 bytes
Database Buffers          822083584 bytes
Redo Buffers              16904192 bytes
Database mounted.
SQL> show parameter db_recover

NAME                                TYPE        VALUE
------------------------------------ ----------- ------------------------------
db_recovery_file_dest                string      /opt/app/oracle/fast_recovery_
                                                area
db_recovery_file_dest_size          big integer 4122M
SQL> !
$ df -h /opt/
文件系统              容量  已用  可用 已用%% 挂载点
/dev/mapper/vg_extend-lv_opt
                      493G  423G  45G  91% /opt
$ exit
exit

SQL> alter system set db_recovery_file_dest_size=10G;

System altered.

SQL> alter database open;

Database altered.

SQL> select status from v$instance;

STATUS
------------
OPEN

SQL>

4.制定对应的归档日志删除策略

SQL> set linesize 120
SQL> select * from v$flash_recovery_area_usage;

FILE_TYPE            PERCENT_SPACE_USED PERCENT_SPACE_RECLAIMABLE NUMBER_OF_FILES
-------------------- ------------------ ------------------------- ---------------
CONTROL FILE                          0                        0              0
REDO LOG                              0                        0              0
ARCHIVED LOG                      41.54                        0            133
BACKUP PIECE                          0                        0              0
IMAGE COPY                            0                        0              0
FLASHBACK LOG                        0                        0              0
FOREIGN ARCHIVED LOG                  0                        0              0
rows selected.
可以rman直接删除不再需要的归档，比如删除1天之前的归档
RMAN> delete noprompt archivelog until time 'sysdate-1';
此处省略删除归档的输出..
RMAN> exit

Recovery Manager complete.

SQL> select * from v$flash_recovery_area_usage;

FILE_TYPE            PERCENT_SPACE_USED PERCENT_SPACE_RECLAIMABLE NUMBER_OF_FILES
-------------------- ------------------ ------------------------- ---------------
CONTROL FILE                          0                        0              0
REDO LOG                              0                        0              0
ARCHIVED LOG                        .35                        0              1
BACKUP PIECE                          0                        0              0
IMAGE COPY                            0                        0              0
FLASHBACK LOG                        0                        0              0
FOREIGN ARCHIVED LOG                  0                        0              0
rows selected.

当然还可以设置crontab定时每天删除1天前的归档：

$ crontab -l
04 * * * /opt/shell/del_arch.sh
$ more /opt/shell/del_arch.sh
#!/bin/bash
export ORACLE_BASE=/opt/app/oracle
export ORACLE_HOME=/opt/app/oracle/product/11.2.0.4/dbhome_1
export ORACLE_SID=JJFAB
export NLS_LANG="simplified chinese_china.ZHS16GBK"
export NLS_DATE_FORMAT="YYYY -MM-DD HH24:Mi:SS"
export PATH=.:/opt/app/oracle/product/11.2.0.4/dbhome_1:/usr/sbin:$ORACLE_HOME/bin:$PATH
export LD_LIBRARY_PATH=$ORACLE_HOME/lib

rman target / log=/opt/backup/del_arch.log <<EOF
list archivelog all;
delete noprompt archivelog until time 'sysdate-1';
show all;
list archivelog all;
EOF