--linux环境10g的rac库
--rman RAC备份出错如下（部分）
Starting Control File and SPFILE Autobackup at 13-SEP-12
piece handle=+DATA/racdb/autobackup/2012_09_13/s_793904799.271.793904805 comment=NONE
Finished Control File and SPFILE Autobackup at 13-SEP-12
sql statement: alter system archive log current
Starting backup at 13-SEP-12
current log archived
released channel: c1
released channel: c2
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of backup command at 09/13/2012 17:07:11
RMAN-06059: expected archived log not found, lost of archived log compromises recoverability
ORA-19625: error identifying file /opt/ora10g/racdb1_arch/1_2_792374929.dbf
ORA-27037: unable to obtain file status
Linux Error: 2: No such file or directory
Additional information: 3
--分析，这个原因有3
1、因为手动删除过归档日志，而未更新相关记录导致rman无法确定其文件状态
解决办法：crosscheck后删除过期文件即可
2、对于rac中的归档的备份，需要确定rman可以正常访问到每个节点的归档日志（可以人为将归档拷贝到相关节点目录）
3、对于rac
解决办法：修改rman配置parallelism和脚本（设置的通道每个节点不够）
参考文档：RMAN Backup Of Archivelog Files On RAC Fails With RMAN-6059 [ID 298119.1]Applies to:
Changes
 Possible modification of setting for RMAN "parallelism".
 
Cause
 This problem can be caused by changing RMAN parallelism to too low a setting
or initially not setting it high enough such that not ALL the node specific channels are allocated.
Solution
 Check your rman parallelism setting:
 
RMAN>show all;
...
 CONFIGURE DEVICE TYPE DISK PARALLELISM 1;
 Make sure parallelism is set to at least the number of channels you're
allocating. For example, if you are issuing this backup command to backup
archivelog files on a two node RAC database:
 
run {
 allocate channel d1 type disk connect'sys/rac@node1';
 allocate channel d2 type disk connect'sys/rac@node2';
 backup archivelog all delete input;
 release channel d1;
 release channel d2;
 }
 then set parallelism to 2:
 
RMAN>CONFIGURE DEVICE TYPE DISK PARALLELISM 2;
 If parallelism is not set at least to the number of channels allocated, not all
channels will be used  and  RMAN will not be able to access all the archivelogs specific to each node.