Ok,?we were able to mount the device on the existing standby node. Mounting on existing standby after upgrading succeeded [root at guard2 ~]# mount -t ocfs2 -o datavolume,nointr /dev/emcpowern1 /orabackup1 Now when we try to mount it on the new standby node, we get the following error Mounting on New standby: [root at guard0 /]# mount -t ocfs2 -o datavolume,nointr /dev/emcpowerf /orabackup mount.ocfs2: Bad magic number in superblock while opening device /dev/emcpowerf Note: /dev/emcpowerf?on the new standby is the same storage as /dev/emcpowern1 on the old standby. Also old standby and new standby are not in the same ocfs2?cluster. Does it have to be in the same cluster? Or can different ocfs2 clusters mount the same device as long as the ocfs2 versions match? ?Regards, Saranya Sivakumar ________________________________ From: Sunil Mushran <sunil.mushran at oracle.com> To: Saranya Sivakumar <sarlavk at yahoo.com> Sent: Wednesday, August 5, 2009 12:05:53 PM Subject: Re: ocfs2 question Always email ocfs2-users. Others can help you too. You have mismatching modules. Not sure how that can happen. Well, if one were to copy the modules around, it could happen. 1. Shutdown o2cb # service o2cb stop Ensure no ocfs2 modules are still loaded in memory. 2. Remove the ocfs2 kernel pacakge. # rpm --erase the ocfs2-2.6.9-42.0.3.ELsmp-1.2.9-1... package. 3. See if an ocfs2 module is still around. # find /lib/modules/`uname -r` -name \*ocfs2\* If so, then someone has been copying modules. Clean that up and reinstall. If you need more help, call oracle support. Sunil Saranya Sivakumar wrote:> Hi > We upgraded ocfs2 to 1.2.9 on the machine, but having errors trying to mount with the new version now. >? /etc/init.d/o2cb status > Module "configfs": Loaded > Filesystem "configfs": Mounted > Module "ocfs2_nodemanager": Loaded > Module "ocfs2_dlm": Not loaded > Module "ocfs2_dlmfs": Not loaded > Filesystem "ocfs2_dlmfs": Not mounted > error while trying to mount >? mount.ocfs2: No such device while mounting /dev/emcpowern1 on /orabackup1. Check 'dmesg' for more information on this error. >? Aug? 5 11:40:18 guard2 modprobe: FATAL: Error inserting ocfs2 (/lib/modules/2.6.9-42.0 > .3.ELsmp/kernel/fs/ocfs2/ocfs2.ko): Unknown symbol in module, or unknown parameter (se > e dmesg) > Aug? 5 11:41:09 guard2 kernel: ocfs2_dlm: disagrees about version of symbol o2net_regi > ster_handler > Aug? 5 11:41:09 guard2 kernel: ocfs2_dlm: Unknown symbol o2net_register_handler > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_print_one_lock > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_errmsg > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlmlock > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_unregister_eviction_cb > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_unregister_domain > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_register_domain > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_register_eviction_cb > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_setup_eviction_cb > Aug? 5 11:41:09 guard2 kernel: ocfs2: disagrees about version of symbol o2net_register > _handler > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol o2net_register_handler > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_errname > Aug? 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlmunlock > >? Please advice >? Saranya > > > > ------------------------------------------------------------------------ > *From:* Sunil Mushran <sunil.mushran at oracle.com> > *To:* Saranya Sivakumar <sarlavk at yahoo.com> > *Cc:* ocfs2-users at oss.oracle.com > *Sent:* Tuesday, August 4, 2009 5:35:21 PM > *Subject:* Re: ocfs2 question > > ocfs2 1.2.3 is 3 years old. Suggest you upgrade that to 1.2.9. > > 1.2.3 and 1.2.9 are not network compatible. The mount will fail. > > Saranya Sivakumar wrote: > > Hi, > > We have a shared backup storage that resides on EMC storage and mounted using ocfs2 1.2.3 on a physical standby database in production. > > We are in the process of adding another physical standby database and need to mount the backup storage on the new physical standby as well, > >? to be able to recover a backup that resides on it. > > But the new standby has ocfs2 1.2.9 installed. > > Eventually we will be removing the current physical standby from the configuration, but for a period of time both physical standby may be using the same backup storage. > > Is it ok to mount the shared storage using different ocfs2 versions on different machines? > > Please advice. > > > > Thanks, > > Sincerely, > > Saranya Sivakumar > > > > > > > > >-------------- next part -------------- An HTML attachment was scrubbed... URL: http://oss.oracle.com/pipermail/ocfs2-users/attachments/20090805/0609c94d/attachment-0001.html
The nodes have to be in the same cluster. Else you will corrupt the fs. Saranya Sivakumar wrote:> Ok, we were able to mount the device on the existing standby node. > > Mounting on existing standby after upgrading succeeded > [root at guard2 ~]# mount -t ocfs2 -o datavolume,nointr /dev/emcpowern1 > /orabackup1 > > Now when we try to mount it on the new standby node, we get the > following error > Mounting on New standby: > [root at guard0 /]# mount -t ocfs2 -o datavolume,nointr /dev/emcpowerf > /orabackup > mount.ocfs2: Bad magic number in superblock while opening device > /dev/emcpowerf > Note: /dev/emcpowerf on the new standby is the same storage as > /dev/emcpowern1 on the old standby. > Also old standby and new standby are not in the same ocfs2 cluster. > Does it have to be in the same cluster? Or can different ocfs2 > clusters mount the same device as long as the ocfs2 versions match? > > Regards, > Saranya Sivakumar > > > ------------------------------------------------------------------------ > *From:* Sunil Mushran <sunil.mushran at oracle.com> > *To:* Saranya Sivakumar <sarlavk at yahoo.com> > *Sent:* Wednesday, August 5, 2009 12:05:53 PM > *Subject:* Re: ocfs2 question > > Always email ocfs2-users. Others can help you too. > > You have mismatching modules. Not sure how that can happen. > Well, if one were to copy the modules around, it could happen. > > 1. Shutdown o2cb > # service o2cb stop > Ensure no ocfs2 modules are still loaded in memory. > > 2. Remove the ocfs2 kernel pacakge. > # rpm --erase the ocfs2-2.6.9-42.0.3.ELsmp-1.2.9-1... package. > > 3. See if an ocfs2 module is still around. > # find /lib/modules/`uname -r` -name \*ocfs2\* > > If so, then someone has been copying modules. Clean that up > and reinstall. > > If you need more help, call oracle support. > > Sunil > > Saranya Sivakumar wrote: > > Hi > > We upgraded ocfs2 to 1.2.9 on the machine, but having errors trying > to mount with the new version now. > > /etc/init.d/o2cb status > > Module "configfs": Loaded > > Filesystem "configfs": Mounted > > Module "ocfs2_nodemanager": Loaded > > Module "ocfs2_dlm": Not loaded > > Module "ocfs2_dlmfs": Not loaded > > Filesystem "ocfs2_dlmfs": Not mounted > > error while trying to mount > > mount.ocfs2: No such device while mounting /dev/emcpowern1 on > /orabackup1. Check 'dmesg' for more information on this error. > > Aug 5 11:40:18 guard2 modprobe: FATAL: Error inserting ocfs2 > (/lib/modules/2.6.9-42.0 > > .3.ELsmp/kernel/fs/ocfs2/ocfs2.ko): Unknown symbol in module, or > unknown parameter (se > > e dmesg) > > Aug 5 11:41:09 guard2 kernel: ocfs2_dlm: disagrees about version of > symbol o2net_regi > > ster_handler > > Aug 5 11:41:09 guard2 kernel: ocfs2_dlm: Unknown symbol > o2net_register_handler > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_print_one_lock > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_errmsg > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlmlock > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol > dlm_unregister_eviction_cb > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol > dlm_unregister_domain > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_register_domain > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol > dlm_register_eviction_cb > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol > dlm_setup_eviction_cb > > Aug 5 11:41:09 guard2 kernel: ocfs2: disagrees about version of > symbol o2net_register > > _handler > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol > o2net_register_handler > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlm_errname > > Aug 5 11:41:09 guard2 kernel: ocfs2: Unknown symbol dlmunlock > > > > Please advice > > Saranya > > > > > > > > ------------------------------------------------------------------------ > > *From:* Sunil Mushran <sunil.mushran at oracle.com > <mailto:sunil.mushran at oracle.com>> > > *To:* Saranya Sivakumar <sarlavk at yahoo.com <mailto:sarlavk at yahoo.com>> > > *Cc:* ocfs2-users at oss.oracle.com <mailto:ocfs2-users at oss.oracle.com> > > *Sent:* Tuesday, August 4, 2009 5:35:21 PM > > *Subject:* Re: ocfs2 question > > > > ocfs2 1.2.3 is 3 years old. Suggest you upgrade that to 1.2.9. > > > > 1.2.3 and 1.2.9 are not network compatible. The mount will fail. > > > > Saranya Sivakumar wrote: > > > Hi, > > > We have a shared backup storage that resides on EMC storage and > mounted using ocfs2 1.2.3 on a physical standby database in production. > > > We are in the process of adding another physical standby database > and need to mount the backup storage on the new physical standby as well, > > > to be able to recover a backup that resides on it. > > > But the new standby has ocfs2 1.2.9 installed. > > > Eventually we will be removing the current physical standby from > the configuration, but for a period of time both physical standby may > be using the same backup storage. > > > Is it ok to mount the shared storage using different ocfs2 > versions on different machines? > > > Please advice. > > > > > > Thanks, > > > Sincerely, > > > Saranya Sivakumar > > > > > > > > > > > > > > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Ocfs2-users mailing list > Ocfs2-users at oss.oracle.com > http://oss.oracle.com/mailman/listinfo/ocfs2-users