1.ocr磁盘简介

健忘问题是由于每个节点都有配置信息的拷贝，修改节点的配置信息不同步引起的。Oracle采用的解决方法就是把这个配置文件放在共享的存储上， 这个文件就是 OCR Disk.整个OCR的信息是树形结构的,主要分为3大分支.分别是SYSTEM,DATABASE和CRS,每个人分支都有许多小分支,只能由root用户修改.ocr每4个小时备份一次,只保留5个文件,当天保留3个文件,上个星期保留一个,上个月也保留一个.

OCR DISK内容包含:

1).节点信息

2).数据库实例,节点,以及其它映射关系

3).rac 资源配置信息(VIP,SERVICES)

4).数据库启动相关进程

5).CRS每三方应用程序信息

2.ocrcheck(检查ocr内容的一至性,查看ocr的状态)

[grid@rac2 ~]$ ocrcheck

Status of Oracle Cluster Registry is as follows :

Version : 3

Total space (kbytes): 262120

Used space (kbytes) : 3140

Available space (kbytes) : 258980

ID : 1119662722

Device/File Name : +OCR

Device/File integrity check succeeded

Device/File not configured

Cluster registry integrity check succeeded

Logical corruption check bypassed due to non-privileged user

[grid@rac2 ~]$

3.ocrdump(查看ocr文件中的内容)

[root@rac1 ~]# ll /home/grid/11.2.0/grid/cdata/scan-cluster/

total 58360

-rw------- 1 root root 7462912 Oct 8 18:44 27255153.ocr

-rw------- 1 root root 7462912 Oct 10 14:44 backup00.ocr

-rw------- 1 root root 7462912 Oct 9 17:58 backup01.ocr

-rw------- 1 root root 7462912 Oct 9 13:58 backup02.ocr

-rw------- 1 root root 7462912 Oct 10 14:45 day_.ocr

-rw------- 1 root root 7462912 Oct 9 13:58 day.ocr

-rw------- 1 root root 7462912 Oct 9 17:58 week_.ocr

-rw------- 1 root root 7421952 Sep 26 14:48 week.ocr

[root@rac1 ~]# /home/grid/11.2.0/grid/bin/ocrdump -backupfile /home/grid/11.2.0/grid/cdata/scan-cluster/backup00.ocr

[root@rac1 ~]# ll OCRDUMPFILE -此时当前目生成OCRDUMPFILE文件

-rw------- 1 root root 220258 Oct 10 16:32 OCRDUMPFILE

[root@rac1 ~]# more OCRDUMPFILE

10/10/2016 16:32:25

/home/grid/11.2.0/grid/cdata/scan-cluster/backup00.ocr

/home/grid/11.2.0/grid/bin/ocrdump.bin -backupfile /home/grid/11.2.0/grid/cdata/scan-cluster/backup00.ocr

[SYSTEM]

UNDEF :

SECURITY : {USER_PERMISSION : PROCR_ALL_ACCESS,GROUP_PERMISSION : PROCR_READ,OTHER_PERMISSION : PROCR_READ,USER_NAME : root,GROUP_NAME : root}

[root@rac1 ~]#

4.ocrconfig(维护ocr磁盘)

[root@rac1 ~]# /home/grid/11.2.0/grid/bin/ocrconfig -showbackup

rac1 2016/10/10 14:45:03 /home/grid/11.2.0/grid/cdata/scan-cluster/backup00.ocr

rac1 2016/10/09 17:58:37 /home/grid/11.2.0/grid/cdata/scan-cluster/backup01.ocr

rac1 2016/10/09 13:58:33 /home/grid/11.2.0/grid/cdata/scan-cluster/backup02.ocr

rac1 2016/10/09 13:58:33 /home/grid/11.2.0/grid/cdata/scan-cluster/day.ocr

rac1 2016/09/26 14:48:52 /home/grid/11.2.0/grid/cdata/scan-cluster/week.ocr

[root@rac1 ~]# /home/grid/11.2.0/grid/bin/ocrconfig -manualbackup --启用手动备份

rac1 2016/10/10 17:44:03 /home/grid/11.2.0/grid/cdata/scan-cluster/backup_20161010_174403.ocr

[root@rac1 ~]# /home/grid/11.2.0/grid/bin/ocrconfig -restore /home/grid/11.2.0/grid/cdata/scan-cluster/backup_20161010_174403.ocr --恢复ocr信息

[root@rac1 ~]#

5.手动备份ocr磁盘

[root@rac1 ~]# /home/grid/11.2.0/grid/bin/ocrconfig -export /tmp/ocr-20161010.back

[root@rac1 ~]# ll /tmp/ocr-20161010.back

-rw------- 1 root root 125185 Oct 10 17:49 /tmp/ocr-20161010.back

[root@rac1 ~]# /home/grid/11.2.0/grid/bin/ocrconfig -import /tmp/ocr-20161010.back --导入时必须停止crs服务

6.vote disk磁盘简介

Voting Disk这个文件主要用于记录节点成员状态，在出现脑裂时，决定那个Partion获得控制权，其他的 Partion必须从集群中剔除。在安装RAC时也会提示指定这个位置。安装完成后可以通过crsctl query css votedisk命令来查看Voting Disk位置.oracle rac 10G可以用dd工具备份表决磁盘,oracle rac 11g的表决磁盘和ocr磁盘是在一起的，不能用dd工具备份.

1).备份表决磁盘

dd if=/dev/raw/raw3 of=/tmp/votedisk.back bs=1024M count=10

2).恢复表决磁盘

dd if=/tmp/votedisk.backof=/dev/raw/raw3 bs=1024M count=10