For data files,it is ok to backup on any RAC node as they are shared by RAC instances;
If parameter files are not in the same place, it should be paid attention to.
If control files are not in the same place, it aslo should be paid attention to.
The Archive logs should be mergered to the same place as, by default, they are placed in the different RAC instances.
The database recovering needs to read all the archive log on all the instances;then firstly we need to make sure how to centralze all the archive logs in the same repository.
show parameter instance_name/spfile
select name from v$datafile;
select name from v$controlfile;
select name, thread# from v$archived_log
select name, thread# from v$archived_dest
Here are the steps to centralize the archive log for all the rac nodes
alter system set log_archiveLdest=' ' scope=both sid='rac1';//this parameter is for local archive log register
alter system set log_archiveLdest1='location=/archive1' scope=both sid='rac1';//this kind of arvhive log parameter is intended for network archive log collection.
alter system set log_archiveLdest2='location=/archive1' scope=both sid='rac1';
The archiveLdest will be conflicted with archiveLdest{number} and one of those needs to be null;
RAC环境下归档日志文件集中存储方法
1. 集中放到共享存储的ASM 磁盘组中
2. 集中放到共享存储上的集群文件系统中
3. 通过NFS方式
4. CROSS-INSTNACE方式select inst_id, name, value from gv$parameter where name='log_archive_dest_1' or name='log_archive_dest_2' or name like '%standby%'select insta_id, dest_name errors,status,process from gv$archive_dest where dest_name='log_arvhice_dest_2'