I don't understand the point on having an OCFS2 volume that can be mounted rw on only one node. In that case, save yourself all the fuzz and hassle and install OCFS2 on just one node.
Regards,
-----Original Message-----
From: Alexei_Roudnev [mailto:[EMAIL PROTECTED]]
Sent: Fri 8/11/2006 8:21 PM
To: Robinson Maureira Castillo; Milind Dumbare; tao.ma
Cc: [email protected]
Subject: Re: [Ocfs2-users] Can not mount
RE: [Ocfs2-users] Can not mountNo, you CAN run OCFSv2 on DRBD. DRBD [rovides you _shared_ storage_.
But, you will have numerous time constrains and I dont think tat such combination can be used for anything except
learning.
(I saw manual, how to instal OCFSv2 on DRBD.)
----- Original Message -----
From: Robinson Maureira Castillo
To: Milind Dumbare ; tao.ma
Cc: [email protected]
Sent: Friday, August 11, 2006 2:56 PM
Subject: RE: [Ocfs2-users] Can not mount
I guess you can't combine DRBD and OCFS2.
DRBD works in the following way, you create a local partition on both nodes, configure DRBD, when you do that it
creates a special device /dev/drbdX or /dev/nbdX, when you start DRBD on the nodes, one becomes the master and the other
becomes slave. You write to the master block device (/dev/drbdX) then the data is copied over the slave, and both write
it to their undelying local storage (/dev/hda9 in your case)
OCFS2 provides access to all nodes concurrently, so it's possible to format an DRBD block device as OCFS2 volume, but
just one node can write to it, the master, in that case (just one node can write to the device) OCFS2 is pointless.
If you want to try OCFS2, you need a shared storage, like shared SCSI bus, access to an SAN volume, iSCSI, etc. that
can be seen as local storage on all nodes.
Regards,
-----Original Message-----
From: [EMAIL PROTECTED] on behalf of Milind Dumbare
Sent: Fri 8/11/2006 4:21 AM
To: tao.ma
Cc: [email protected]
Subject: Re: [Ocfs2-users] Can not mount
Ohh, Do you have any idea how DRBD will work here?
On Fri, 2006-08-11 at 16:11 +0800, tao.ma wrote:
> ocfs2 requires a shared disk. As in, all nodes must be able to concurrently
> read/write to the device.
> So your second node can't see "/dev/hda9" of your first node. That is
> the problem.
>
> Milind Dumbare wrote:
>
> >HI all,
> > I have two nodes in cluster, my /etc/ocfs2/cluster.conf as follows
> >==============================================
> > cluster:
> > node_count = 2
> > name = mili
> > node:
> > ip_port = 7777
> > ip_address = 192.168.1.23
> > number = 7
> > name = panini
> > cluster = mili
> > node:
> > ip_port = 7777
> > ip_address = 192.168.1.22
> > number = 6
> > name = xenon
> > cluster = mili
> >===============================================
> >I did
> > #/etc/init.d/o2cb load
> > &
> > #/etc/init.d/o2cb start mili
> >
> >on both nodes it was successful.
> >
> >Then I did
> > # mkfs.ocfs2 -b 4k -C 32K -L "drbd" -N 2 /dev/hda9
> >on panini it was successful too
> >then
> > #mount -vt ocfs2 -L "test" /mnt/store/
> >on panini, success again.
> >
> >now when I try to mount "test" on xenon (other node) its saying
> > "no such partition"
> >
> >Please help me out.
> >
> >
> >
>
>
>
--
-Milind
"There is no place like 127.0.0.1"
_______________________________________________
Ocfs2-users mailing list
[email protected]
http://oss.oracle.com/mailman/listinfo/ocfs2-users
------------------------------------------------------------------------------
_______________________________________________
Ocfs2-users mailing list
[email protected]
http://oss.oracle.com/mailman/listinfo/ocfs2-users
_______________________________________________ Ocfs2-users mailing list [email protected] http://oss.oracle.com/mailman/listinfo/ocfs2-users
