今天遇到一个RMAN备份的错误,报错信息如下:
Starting backup at 2007-09-03:08:19:42
user interrupt received
error occurred in source file: krmk.pc, line: 7635
text of failing SQL statement: alter system archive log current
ORACLE error from target database:
ORA-01013: user requested cancel of current operation
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of backup command at 09/03/2007 08:24:47
RMAN-03014: implicit resync of recovery catalog failed
RMAN-03099: job cancelled at user request
Recovery Manager complete.
从以上信息可以看出RMAN在执行alter system archive log current命令的时候报错,导致备份没有成功.
查看alert日志,发现如下信息:
Mon Sep 3 08:31:23 2007
ARCH: Evaluating archive log 11 thread 1 sequence 2185
ARCH: Beginning to archive log 11 thread 1 sequence 2185
Creating archive destination LOG_ARCHIVE_DEST_2: '/export/home/oracle/dev/MSP/log_archive_dest_2/1_2185.dbf'
Mon Sep 3 08:31:24 2007
ARC0: Evaluating archive log 11 thread 1 sequence 2185
ARC0: Unable to archive log 11 thread 1 sequence 2185
Log actively being archived by another process
Mon Sep 3 08:31:24 2007
Creating archive destination LOG_ARCHIVE_DEST_1: '/export/home/oracle/dev/MSP/log_archive_dest_1/1_2185.dbf'
ARCH: Completed archiving log 11 thread 1 sequence 2185
Log actively being archived by another process说明归档进程可能会有2个或者2个以上,如果其中一个正在归档某个日志，而另一个进程来尝试归档该日志的话，则会出现这个提示，这个问题一般在使用rman执行backup加选项plus archivelog 时候出现。要解决这个问题你可以在如何能提高归档速度方面考虑，比如日志文件和归档目的地的写是否存在IO上的可能的提高（比如归档目的地是raid5换做raid10等）,或者某个归档进程出现异常或者归档目的地有问题造成归档进程阻塞而没有及时完成.
结合以上信息，可以断定备份失败和Log actively being archived by another process有关系。原因是当数据库正在执行alter system archive log current的时候，RMAN备份也刚好发起了一个归档日志的操作，这样就造成了两个归档进程之间的竞争,从而导致备份失败。
关于这个问题产生的详细原因可以参考eygle的一篇文章:
http://www.eygle.com/archives/2007/06/log_archive_process_more.html