请问个RAC环境下的RMAN备份问题

keymanlu

有个RAC集群环境，db1和db2

各自机器上都有自己的归档日志/arch01  /arch02的FS用来放各自的归档。

当在db1上启动备份的时候，为什么会去找db2上的归档日志呢？
因为我在每天的备份日志里，都看到有下面这样的错误：

Starting backup at 26-APR-10
current log archived
released channel: c1
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of backup command at 04/26/2010 01:00:13
RMAN-06059: expected archived log not found, lost of archived log compromises recoverability
ORA-19625: error identifying file /arch02/esdb/2_168_706025174.dbf
ORA-27037: unable to obtain file status
IBM AIX RISC System/6000 Error: 2: No such file or directory
Additional information: 3

RMAN>




我的归档脚本如下，是仿照网上的朋友的来改动的：
#!/bin/ksh
#set env
export ORACLE_BASE=/oracle/products
export ORACLE_HOME=$ORACLE_BASE/10.2/db
export ORACLE_SID=esdb1
export NLS_LANG=AMERICAN_AMERICA.ZHS16GBK
export PATH=$ORACLE_HOME/binPATH
DATE=`date +%w`
DATE_2=`date +%Y%m%d`
BACKUP_PATH=/arch01/rman_bak
LEVEL=$@
TARGET_SID=db1
RMAN_SID=db1.db1
BIN=$ORACLE_HOME/bin
# Delete the data backuped last time
rm -rf $BACKUP_PATH/data/$DATE/*

if [ $# != 1 ]; then
echo "usage: rman_bak.sh n
where n is the rman backup level(Only 0,1,2 is permitted)."
exit 1
fi


if [ $@ -ne 0 -a $@ -ne 1 -a $@ -ne 2 ]; then
echo "usage: rman_bak.sh n
where n is the rman backup level(Only 0,1,2 is permitted)."
exit 2
fi


echo "-----------------------------start-----------------------------";date
if [ $LEVEL = 0 ]; then
$BIN/rman log $BACKUP_PATH/log/level.$TARGET_SID.$LEVEL.$DATE_2.log <<EOF
connect target sys/oracle;
#connect catalog rman/rman@$RMAN_SID;
#resync catalog;
run{
allocate channel c1 type disk ;
crosscheck backupset of archivelog all ;
backup filesperset 3 format '$BACKUP_PATH/data/$DATE/arch.%d.live.$LEVEL.%t'(archivelog from time 'sysdate-7' all delete input) ;
delete noprompt expired backupset of archivelog all ;
release channel c1 ;
}

run{
allocate channel c2 type disk ;
crosscheck backupset of database ;
backup incremental level $LEVEL filesperset 3 format '$BACKUP_PATH/data/$DATE/data.%d.live.$LEVEL.%t'(database include current controlfile) ;
delete noprompt expired backupset of database ;
delete noprompt obsolete ;
release channel c2 ;
}
exit;
EOF

else
$BIN/rman log $BACKUP_PATH/log/level.$TARGET_SID.$LEVEL.$DATE_2.log <<EOF
connect target sys/oracle;
#connect catalog rman/rman@$RMAN_SID;
#resync catalog;
run{
allocate channel c1 type disk ;
crosscheck backupset of archivelog all ;
backup filesperset 3 format '$BACKUP_PATH/data/$DATE/arch.%d.live.$LEVEL.%t' (archivelog from time 'sysdate-1' all) ;
delete noprompt expired backupset of archivelog all ;
release channel c1 ;
}

run{
allocate channel c2 type disk ;
crosscheck backupset of database ;
backup incremental level $LEVEL filesperset 3 format '$BACKUP_PATH/data/$DATE/data.%d.live.$LEVEL.%t' (database include current controlfile) ;
delete noprompt expired backupset of database ;
delete noprompt obsolete ;
release channel c2 ;
}

exit;
EOF

fi
#$BIN/exp rman/rman@$RMAN_SID file=$BACKUP_PATH/data/$DATE/live.rman.$DATE_2.dmp log=$BACKUP_PATH/log/$RMAN_SID.rman.$DATE_2.log
echo "------------------------------end------------------------------";date

renxiao2003

1. RMAN-03002: failure of backup command at 04/26/2010 01:00:13
   
2. RMAN-06059: expected archived log not found, lost of archived log compromises recoverability
   
3. ORA-19625: error identifying file /arch02/esdb/2_168_706025174.dbf
   
4. ORA-27037: unable to obtain file status
   
5. IBM AIX RISC System/6000 Error: 2: No such file or directory
   
6. Additional information: 3
   

复制代码

从这里看。是不是dbf文件不存在啊。

keymanlu

从这里看。是不是dbf文件不存在啊。
renxiao2003 发表于 2010-04-26 11:20

    是不存在的，因为/arch01 和/arch02 分别是在两个节点服务器db1 db2服务器上的。
我的疑问就是说为什么在节点db1上的rman会去找节点db2上的归档日志呢？

zyhwork778

可能是2台机器间的nfs参数不对
options     = cio,rw,bg,hard,intr,rsize=32768,wsize=32768,proto=tcp,noac,vers=3,timeo=600
根据环境不同，参数可能也不同

andrew_young

rac的归档不在一个共享目录下的，rman需要起两个channel读各自节点的归档

allocate channel c1 device type disk connect'sys/***@node1_sid';
allocate channel c2 device type disk connect'sys/***@node2_sid';