Oracle package failed to switch

[Denn]

I am having problem switching back the oracle package to its primary node but it works fine on the secondary node and It was also working fine before the server was moved to different location.
The info I got from the cntl log file showed that the switching was hung during the filesystem mounting process. Would appreciate that someone could shed some light.

~~
Activated volume group in Exclusive Mode.
Volume group "/dev/vgcomdata" has been successfully changed.
Node "prod": Checking filesystems:
/dev/vgcom/lvol1
/dev/vgcomdata/lvol1
/dev/vgcomdata/lvol2
/dev/vgcom/rlvol1:file system is clean - log replay is not required
/dev/vgcomdata/rlvol1:file system is clean - log replay is not required
/dev/vgcomdata/rlvol2:file system is clean - log replay is not required
Node "prod": Mounting /dev/vgcom/lvol1 at /u01

Thanks

[freddy_21]

Hello,

could you attach cluster log file on both server. usually at /etc/cmcluster//.cntl.log

from that file, can found what is your problem.

thanks
freddy

[Senthil Kumar .A_1]

Hi,

What happens when you mount manually.

for instance...

vgimport /dev/vgcomdata .....
vgchange -a y /dev/vgcomdata
mount /dev/vgcom/lvol1 /u01

Do you get any errors...?

SK

[Senthil Kumar .A_1]

oops...

I meant


What happens when you mount manually.

for instance...

vgimport /dev/vgcom .....
vgchange -a y /dev/vgcom
mount /dev/vgcom/lvol1 /u01

Do you get any errors...?

SK

[Andrew Young_2]

Hi SK

The vgchange command should be

vgchange -a e /dev/vgcom

to lock it in exclusive mode.

Is there anything in either /var/adm/syslog/syslog.log?

Regards

AY

[Denn]

Hi,

Thanks for the suggestion. Attached are the log files. As they are the production servers hence, I cannot simply test it out without the downtime approval. I hope that the log files could give some hints.

[Denn]

Performed clean stop on application and tested with manual vg activation and mount.