作者归档：惜分飞

RAC主机相差超过10分钟导致crs无法启动

发表于 2024 年 6 月 11 日 由 惜分飞

客户反馈有一套19c 2节点rac，断电之后，一个节点数据库无法正常启动,通过crsctl命令查看发现crs进程没有正常启动

[root@xifenf1 ~]# /u01/app/19.0/grid/bin/crsctl status res -t -init
--------------------------------------------------------------------------------
Name           Target  State        Server                   State details
--------------------------------------------------------------------------------
Cluster Resources
--------------------------------------------------------------------------------
ora.asm
      1        ONLINE  OFFLINE                               STABLE
ora.cluster_interconnect.haip
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.crf
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.crsd
      1        ONLINE  OFFLINE                               STABLE
ora.cssd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.cssdmonitor
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.ctssd
      1        ONLINE  OFFLINE                               STABLE
ora.diskmon
      1        OFFLINE OFFLINE                               STABLE
ora.evmd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.gipcd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.gpnpd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.mdnsd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.storage
      1        ONLINE  ONLINE       xifenf1                  STABLE
--------------------------------------------------------------------------------

查看crs的alert日志发现集群时间间隔超过600s，无法启动csst进程

2024-06-11 17:33:09.953 [OCSSD(5020)]CRS-1605: CSSD voting file is online: /dev/asm_ocr5; 
 details in /u01/app/grid/diag/crs/xifenf1/crs/trace/ocssd.trc.
2024-06-11 17:33:09.956 [OCSSD(5020)]CRS-1605: CSSD voting file is online: /dev/asm_ocr1; 
 details in /u01/app/grid/diag/crs/xifenf1/crs/trace/ocssd.trc.
2024-06-11 17:33:10.024 [OCSSD(5020)]CRS-1605: CSSD voting file is online: /dev/asm_ocr2; 
 details in /u01/app/grid/diag/crs/xifenf1/crs/trace/ocssd.trc.
2024-06-11 17:33:10.031 [OCSSD(5020)]CRS-1605: CSSD voting file is online: /dev/asm_ocr4; 
 details in /u01/app/grid/diag/crs/xifenf1/crs/trace/ocssd.trc.
2024-06-11 17:33:10.040 [OCSSD(5020)]CRS-1605: CSSD voting file is online: /dev/asm_ocr3; 
 details in /u01/app/grid/diag/crs/xifenf1/crs/trace/ocssd.trc.
2024-06-11 17:33:11.900 [OCSSD(5020)]CRS-1601: CSSD Reconfiguration complete. Active nodes are xifenf1 xifenf2 .
2024-06-11 17:33:13.344 [OCSSD(5020)]CRS-1720: Cluster Synchronization Services daemon (CSSD) is ready for operation.
2024-06-11 17:33:13.809 [OCTSSD(5488)]CRS-8500: Oracle Clusterware OCTSSD process is starting with operating system process ID 5488
2024-06-11 17:33:16.017 [OCTSSD(5488)]CRS-2407: The new Cluster Time Synchronization Service reference node is host xifenf2.
2024-06-11 17:33:16.018 [OCTSSD(5488)]CRS-2401: The Cluster Time Synchronization Service started on host xifenf1.
2024-06-11 17:33:16.105 [OCTSSD(5488)]CRS-2419: The clock on host xifenf1 differs from mean cluster time by 1031504618 microseconds. 
  The Cluster Time Synchronization Service will not perform time synchronization 
  because the time difference is beyond the permissible offset of 600 seconds. 
  Details in /u01/app/grid/diag/crs/xifenf1/crs/trace/octssd.trc.
2024-06-11 17:33:16.579 [OCTSSD(5488)]CRS-2402: The Cluster Time Synchronization Service aborted on host xifenf1. 
  Details at (:ctsselect_mstm4:) in /u01/app/grid/diag/crs/xifenf1/crs/trace/octssd.trc.

查看主机时间

[grid@xifenf1 ~]$ date ;ssh xifenf2 date
Tue Jun 11 17:54:09 CST 2024
Tue Jun 11 18:04:34 CST 2024

修改主机时间

[root@xifenf1 ~]# date -s "20240611 18:06:00"
Tue Jun 11 18:06:00 CST 2024
[root@xifenf1 ~]# su - grid
Last login: Tue Jun 11 17:37:53 CST 2024 on pts/0
[grid@xifenf1 ~]$ date ;ssh xifenf2 date
Tue Jun 11 18:06:09 CST 2024
Tue Jun 11 18:05:34 CST 2024

重启crs

[root@xifenf1 ~]# /u01/app/19.0/grid/bin/crsctl stop crs -f
CRS-2791: Starting shutdown of Oracle High Availability Services-managed resources on 'xifenf1'
CRS-2673: Attempting to stop 'ora.storage' on 'xifenf1'
CRS-2673: Attempting to stop 'ora.mdnsd' on 'xifenf1'
CRS-2673: Attempting to stop 'ora.crf' on 'xifenf1'
CRS-2677: Stop of 'ora.storage' on 'xifenf1' succeeded
CRS-2673: Attempting to stop 'ora.evmd' on 'xifenf1'
CRS-2673: Attempting to stop 'ora.cluster_interconnect.haip' on 'xifenf1'
CRS-2677: Stop of 'ora.mdnsd' on 'xifenf1' succeeded
CRS-2677: Stop of 'ora.crf' on 'xifenf1' succeeded
CRS-2677: Stop of 'ora.evmd' on 'xifenf1' succeeded
CRS-2677: Stop of 'ora.cluster_interconnect.haip' on 'xifenf1' succeeded
CRS-2673: Attempting to stop 'ora.cssd' on 'xifenf1'
CRS-2677: Stop of 'ora.cssd' on 'xifenf1' succeeded
CRS-2673: Attempting to stop 'ora.gpnpd' on 'xifenf1'
CRS-2673: Attempting to stop 'ora.gipcd' on 'xifenf1'
CRS-2677: Stop of 'ora.gpnpd' on 'xifenf1' succeeded
CRS-2677: Stop of 'ora.gipcd' on 'xifenf1' succeeded
CRS-2793: Shutdown of Oracle High Availability Services-managed resources on 'xifenf1' has completed
CRS-4133: Oracle High Availability Services has been stopped.
[root@xifenf1 ~]# /u01/app/19.0/grid/bin/crsctl start crs
CRS-4123: Oracle High Availability Services has been started.
[root@xifenf1 ~]# /u01/app/19.0/grid/bin/crsctl status res -t -init
--------------------------------------------------------------------------------
Name           Target  State        Server                   State details
--------------------------------------------------------------------------------
Cluster Resources
--------------------------------------------------------------------------------
ora.asm
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.cluster_interconnect.haip
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.crf
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.crsd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.cssd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.cssdmonitor
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.ctssd
      1        ONLINE  ONLINE       xifenf1                  ACTIVE:35600,STABLE
ora.diskmon
      1        OFFLINE OFFLINE                               STABLE
ora.evmd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.gipcd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.gpnpd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.mdnsd
      1        ONLINE  ONLINE       xifenf1                  STABLE
ora.storage
      1        ONLINE  ONLINE       xifenf1                  STABLE
--------------------------------------------------------------------------------
发表在 Oracle RAC | 标签为 | 评论关闭

O/S-Error: (OS 23) 数据错误(循环冗余检查)—故障处理

发表于 2024 年 6 月 10 日 由 惜分飞

有客户由于磁盘坏道导致数据文件访问报ORA-27070 OSD-04016 O/S-Error等相关错误
OSD-04016

rman 尝试读取88号文件

RMAN> backup datafile 88 format 'e:\rman\%d_%T_%I.%s%p';

启动 backup 于 05-6月 -24
使用目标数据库控制文件替代恢复目录
分配的通道: ORA_DISK_1
通道 ORA_DISK_1: SID=2246 设备类型=DISK
通道 ORA_DISK_1: 正在启动全部数据文件备份集
通道 ORA_DISK_1: 正在指定备份集内的数据文件
输入数据文件: 文件号=00088 名称=E:\APP\ADMINISTRATOR\ORADATA\XFF\XIFENFEI.ORA
通道 ORA_DISK_1: 正在启动段 1 于 05-6月 -24
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: backup 命令 (ORA_DISK_1 通道上, 在 06/05/2024 18:33:43 上) 失败
ORA-19501: 文件 "E:\APP\ADMINISTRATOR\ORADATA\XFF\XIFENFEI.ORA", 块编号 322944 (块大小=8192) 上出现读取错误
ORA-27070: 异步读取/写入失败
OSD-04016: 异步 I/O 请求排队时出错。
O/S-Error: (OS 23) 数据错误(循环冗余检查)。

检查系统日志,发现有报:设备 \Device\Harddisk0\DR0 有一个不正确的区块
disk-error

到目前为止基本上可以判断是文件系统或者磁盘层面出现坏道导致该问题（磁盘坏道概率更大）,使用工具对损坏数据文件进行强制拷贝,提示少量扇区数据无法拷贝
force-copy
通过dbv检查恢复出来文件效果

E:\check_db>DBV FILE=D:/XIFENFEI.ORA

DBVERIFY: Release 11.2.0.3.0 - Production on 星期日 6月 9 22:17:28 2024

Copyright (c) 1982, 2011, Oracle and/or its affiliates.  All rights reserved.

DBVERIFY - 开始验证: FILE = D:\XIFENFEI.ORA
页 323055 标记为损坏
Corrupt block relative dba: 0x1604edef (file 88, block 323055)
Bad header found during dbv:
Data in bad block:
 type: 229 format: 5 rdba: 0xe5e5e5e5
 last change scn: 0xe5e5.e5e5e5e5 seq: 0xe5 flg: 0xe5
 spare1: 0xe5 spare2: 0xe5 spare3: 0xe5e5
 consistency value in tail: 0x4c390601
 check value in block header: 0xe5e5
 computed block checksum: 0x5003



DBVERIFY - 验证完成

检查的页总数: 524288
处理的页总数 (数据): 204510
失败的页总数 (数据): 0
处理的页总数 (索引): 127485
失败的页总数 (索引): 0
处理的页总数 (其他): 3030
处理的总页数 (段)  : 0
失败的总页数 (段)  : 0
空的页总数: 189262
标记为损坏的总页数: 1
流入的页总数: 0
加密的总页数        : 0
最高块 SCN            : 184063522 (3470.184063522)

运气不错，就一个数据库block异常,通过dba_extents查询坏块所属对象，运气不太好是一个表数据

SQL> SELECT OWNER, SEGMENT_NAME, SEGMENT_TYPE, TABLESPACE_NAME, A.PARTITION_NAME
  2    FROM DBA_EXTENTS A
  3   WHERE FILE_ID = &FILE_ID
  4     AND &BLOCK_ID BETWEEN BLOCK_ID AND BLOCK_ID + BLOCKS - 1;
输入 file_id 的值:  88
原值    3:  WHERE FILE_ID = &FILE_ID
新值    3:  WHERE FILE_ID = 88
输入 block_id 的值:  323055
原值    4:    AND &BLOCK_ID BETWEEN BLOCK_ID AND BLOCK_ID + BLOCKS - 1
新值    4:    AND 323055 BETWEEN BLOCK_ID AND BLOCK_ID + BLOCKS - 1

OWNER
------------------------------
SEGMENT_NAME
--------------------------------------------------------------------------------
SEGMENT_TYPE       TABLESPACE_NAME                PARTITION_NAME
------------------ ------------------------------ ------------------------------
XFFUSER
XFFTABLE
TABLE              XFFTBS_TAB2

设置跳过坏块导出数据，然后重命名原表导入数据,完成本次恢复，以前有过类似恢复:

一次侥幸的OSD-04016 O/S-Error异常恢复
O/S-Error: (OS 23) 数据错误(循环冗余检查) 数据库恢复

发表在 Oracle备份恢复 | 标签为 , , | 评论关闭

异常断电数据库恢复-从ORA-600 2131到ORA-08102: 未找到索引关键字, 对象号 39

发表于 2024 年 6 月 5 日 由 惜分飞

数据库启动报ORA-600 2131，以前遇到过类似问题:ORA-600 2131故障处理

SQL> alter database mount;
alter database mount
*
第 1 行出现错误:
ORA-00600: ??????, ??: [2131], [9], [8], [], [], [], [], [], [], [], [], []

Tue Jun 04 14:12:18 2024
RECO started with pid=15, OS id=3244 
Tue Jun 04 14:12:18 2024
MMON started with pid=16, OS id=3256 
Tue Jun 04 14:12:18 2024
MMNL started with pid=17, OS id=3432 
starting up 1 dispatcher(s) for network address '(ADDRESS=(PARTIAL=YES)(PROTOCOL=TCP))'...
starting up 1 shared server(s) ...
ORACLE_BASE from environment = E:\app\Administrator
Tue Jun 04 14:12:22 2024
alter database mount exclusive
Errors in file E:\APP\ADMINISTRATOR\diag\rdbms\XFF\XFF\trace\XFF_ora_2536.trc  (incident=427583):
ORA-00600: ??????, ??: [2131], [9], [8], [], [], [], [], [], [], [], [], []
Tue Jun 04 14:12:28 2024
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
ORA-600 signalled during: alter database mount exclusive...

重建ctl，然后重试recover 数据库，报ORA-600 kdourp_inorder2和ORA-600 3020错误，这些错误本质都是由于redo信息和block信息不匹配导致

SQL> recover datafile 1;
ORA-00283: 恢复会话因错误而取消
ORA-10562: Error occurred while applying redo to data block (file# 1, block# 74805)
ORA-10564: tablespace SYSTEM
ORA-01110: 数据文件 1: 'E:\ORADATA\XFF\SYSTEM01.DBF'
ORA-10561: block type 'TRANSACTION MANAGED DATA BLOCK', data object# 8
ORA-00600: 内部错误代码, 参数: [kdourp_inorder2], [16], [3], [0], [108], [], [], [], [], [], [], []


SQL> recover datafile 7;
ORA-00283: 恢复会话因错误而取消
ORA-00600: 内部错误代码, 参数: [3020], [7], [385], [29360513], [], [], [], [], [], [], [], []
ORA-10567: Redo is inconsistent with data block (file# 7, block# 385, file offset is 3153920 bytes)
ORA-10564: tablespace UNDOTBS2
ORA-01110: 数据文件 7: 'E:\ORADATA\XFF\UNDOTBS2.DBF'
ORA-10560: block type 'KTU UNDO BLOCK'

通过屏蔽一致性,修改文件头scn,强制打开数据库

SQL> recover database until cancel;
ORA-00279: 更改 56782359 (在 06/04/2024 14:00:36 生成) 对于线程 1 是必需的
ORA-00289: 建议: E:\APP\ARCHIVELOG\ARC0000005415_1165094245.0001
ORA-00280: 更改 56782359 (用于线程 1) 在序列 #5415 中


指定日志: {<RET>=suggested | filename | AUTO | CANCEL}
cancel
ORA-01547: 警告: RECOVER 成功但 OPEN RESETLOGS 将出现如下错误
ORA-01194: 文件 1 需要更多的恢复来保持一致性
ORA-01110: 数据文件 1: 'E:\ORADATA\XFF\SYSTEM01.DBF'


ORA-01112: 未启动介质恢复


SQL> alter database open resetlogs;

数据库已更改。

尝试导出数据报ORA-08102,导致数据库无法正常导出

C:\Users\Administrator>expdp "'/ as sysdba'" full=y dumpfile=full_20240604_%U.dmp DIRECTORY=expdp_dir 
logfile=full_20240604.log parallel=2 EXCLUDE=STATISTICS,AUDIT

Export: Release 11.2.0.4.0 - Production on 星期二 6月 4 18:40:26 2024

Copyright (c) 1982, 2011, Oracle and/or its affiliates.  All rights reserved.

连接到: Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
ORA-31626: 作业不存在
ORA-31633: 无法创建主表 "SYS.SYS_EXPORT_FULL_05"
ORA-06512: 在 "SYS.DBMS_SYS_ERROR", line 95
ORA-06512: 在 "SYS.KUPV$FT", line 1038
ORA-08102: 未找到索引关键字, 对象号 39, 文件 1, 块 97540 (2)

obj 39 为OBJ$的I_OBJ4对象报ORA-08102

SQL> select owner,object_name,object_type from dba_objects where object_id=39
  2  /

OWNER                          OBJECT_NAME                    OBJECT_TYPE
------------------------------ ------------------------------ -------------------
SYS                            I_OBJ4                         INDEX

该对象属于bootstrap$中核心对象，无法直接rebuild,参考下面文章处理,然后再尝试导出数据
分享I_OBJ4 ORA-8102故障恢复案例
使用bbed 修复I_OBJ4 index 报ORA-8102错误
bootstrap$核心index(I_OBJ1,I_USER1,I_FILE#_BLOCK#,I_IND1,I_TS#,I_CDEF1等)异常恢复—ORA-00701错误解决

C:\Users\Administrator>expdp "'/ as sysdba'" full=y dumpfile=full_20240604_%U.dmp DIRECTORY=expdp_dir 
logfile=full_20240604.log parallel=2 EXCLUDE=STATISTICS,AUDIT

Export: Release 11.2.0.4.0 - Production on 星期二 6月 4 18:43:47 2024

Copyright (c) 1982, 2011, Oracle and/or its affiliates.  All rights reserved.

连接到: Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
ORA-31626: 作业不存在
ORA-31637: 无法创建作业 SYS_EXPORT_FULL_01 (用户 SYS)
ORA-06512: 在 "SYS.DBMS_SYS_ERROR", line 95
ORA-06512: 在 "SYS.KUPV$FT_INT", line 798
ORA-39080: 无法为数据泵作业创建队列 "KUPC$C_1_20240604184348" 和 "KUPC$S_1_20240604184348"
ORA-06512: 在 "SYS.DBMS_SYS_ERROR", line 95
ORA-06512: 在 "SYS.KUPC$QUE_INT", line 1534
ORA-08102: 未找到索引关键字, 对象号 53, 文件 1, 块 97715 (2)

通过类似方法分析确认为CDEF$的I_CDEF1 index,处理方法和I_OBJ4一样,然后导出数据成功,导入到新库中,在这个迁移过程中遭遇Wrapped 加密的package body无效的问题,具体参见:数据泵迁移Wrapped PLSQL之后报PLS-00753

发表在 Oracle备份恢复 | 标签为 , , , , , , , , | 评论关闭