- 论坛徽章:
- 0
|
我用两台Dell2600系统TurboLinux8.0做Oracle数据库，数据放在一阵列柜上，阵列柜4个硬盘做RAID1+0 文件系统为ext3
两台Dell2600上装有HA5.0，当用HA做切换动作时在HA还没启动脚本之前都会做FSCK磁盘阵列磁盘的动作，在HA的信息栏出现提示
INFO:FSCK is repairing device </dev/sdb1> Please wait....
以下ha_ora_start.sh
#!/bin/sh
# This file: ora_start.sh
# Version: 4.0.1
HAHOME=`cat /etc/init.d/HAHOME`
export HAHOME
out=$HAHOME/bin/APIOUT.x
JOBNAME=$2
# When Another Server are Down, You MUST sleep awhile.
# You can change this value to meet your requirement if need.
if [ "$1" = "anotherdown" ]
then
/bin/sleep 30
fi
DISKDEV=/dev/sdb1
MOUNTPOINT=/oradata
#$HAHOME/bin/dflush $DISKDEV
#fsck -a $DISKDEV
#if test $? -ne 0
# then
# ${out} "[INFO] fsck ${DISKDEV}......"
# fsck -yf $DISKDEV
# fi
mount $DISKDEV $MOUNTPOINT
mount | grep "${DISKDEV} on ${MOUNTPOINT} " >/dev/null 2>&1
if test $? -ne 0
then
${out} "[INFO] Cannot mount ${DISKDEV}."
exit
fi
$HAHOME/bin/ha_ag_oracle.x $JOBNAME 0
if test $? -eq 0
then
${out} "[INFO] The service <${JOBNAME}> has started."
exit
else
${out} "[INFO] Start Oracle server...."
$HAHOME/bin/smgprod_ha abort
$HAHOME/bin/smgprod_ha start
fi
${out} "[INFO] Start shell <$0> finished."
以下是日志文件：
smgprod.log
10/26/06 18:49:01 [INFO]The timeout of check disk status is 120 seconds.
10/26/06 18:49:01 (100)[INFO]HA server started ok.
10/26/06 18:49:05 (701)[INFO]GUI Administration Tool is listening from port <10000>...
10/26/06 18:49:05 (530)[INFO]The service <Ora> execute </opt/raidsysha/bin/ora_stop.sh init Ora 0 &>.
10/26/06 18:49:05 (560)[INFO]Wait </opt/raidsysha/bin/ora_stop.sh> to finish...
10/26/06 18:49:25 (305)[INFO]The path <rs232> timeout over (24) seconds.
10/26/06 18:49:25 (303)[INFO]The path <rs232> is down.
10/26/06 18:49:40 (310)[INFO]The public net is functioning,but all instances of the private net are down or unavailable.
10/26/06 18:49:42 (306)[INFO]The path <rs232> recovers.
10/26/06 18:50:25 (521)[INFO]Fsck is repairing device </dev/sdb1>.Please wait...
10/26/06 18:54:56 (530)[INFO]The service <Ora> execute </opt/raidsysha/bin/ora_start.sh init Ora 3>&2 4>&2 5>&2 6>&2 7>&2 8>&2 9>&2 &>.
mount: /dev/sdb1 already mounted or /oradata busy
mount: according to mtab, /dev/sdb1 is already mounted on /oradata
10/26/06 18:54:56 (560)[INFO]Wait </opt/raidsysha/bin/ora_start.sh> to finish...
Get message queue 10 error: No such file or directory
10/26/06 18:55:39 (530)[INFO]The service <Ora> execute </opt/raidsysha/bin/ha_ag_oracle.x Ora 10 3>&2 4>&2 5>&2 6>&2 7>&2 8>&2 9>&2 &>.
10/26/06 18:55:39 (404)[INFO]Takeover <Ora> :success.
10/26/06 18:55:42 (102)[INFO]HA server inited end.
Exec Shell Program /opt/raidsysha/bin/mail.sh Info "HA server started." )
10/26/06 18:55:47 [INFO]haapi--apiok(Ora).
检查磁盘要发费4---5分钟之久，检完磁盘后就开始加载ora_start.sh脚本，使接管时间过长。
请高人指点。问题出在那部分
谢谢！！！
[ 本帖最后由 qjlpn 于 2006-10-26 22:50 编辑 ] |
|
免费注册
发表于 2006-10-26 22:02
