[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-users] Boot from OCFS2



ïDear Waldirio,

Thank you for your description! (hell yea I can't portuguese but I
understood it ;-)) 

I think you didn't get the point what I would do exactly...
I try to boot from a ocfs2 fs. so the whole system is on a LV which is
formated with ocfs2. I don't know if this possible... but I need a
clustring filesystem because I dont have isci or something else... just
2 boxes which access the same scsi device... 

Greetings
Dominik


Am Donnerstag, den 03.07.2008, 08:05 -0300 schrieb Waldirio ManhÃes
Pinheiro:
>     Good Morning Dominik
>  
>   Let's try explain the implementation of ocfs2 in guest machine (in
> my case RedHat).
>  
>   I have created the Oracle RAC summary and one point of the procedure
> is the implementation of ocfs2. Because of size, I removed all another
> points, keep only ocfs (later I'll put in my web site the full
> document).
>  
>   When you configure using ocfs2console (graphical environment),
> you'll inform the IP address of machines which will access the shared
> file system, at the finish, will be created one ocfs config file
> (/etc/ocfs2/cluster.conf) with this content (in my case)
>  
> ###
> node:
>         ip_port = 7777
>         ip_address = 164.85.216.107
>         number = 0
>         name = cambeba.mydomain.com
>         cluster = ocfs2
> node:
>         ip_port = 7777
>         ip_address = 164.85.216.105
>         number = 1
>         name = cangua.mydomain.com
>         cluster = ocfs2
> 
> cluster:
>         node_count = 2
>         name = ocfs2
> ###
> 
> At the finish of configuration, the tool ocfs2console have the option
> "Propagate Configuration" which will share this configuration with all
> another nodes configured.
> 
> The next step is create a partition (can be using lv for example) and
> format using ocfs2
> 1. Create using lvm (pvcreate/vgcreate/lvcreate)
> 2. Formatting using ocfs2 (mkfs.ocfs2 <path created before for
> lvcreate>)
> 3. Mount automatically the partition by fstab (include in fstab like
> below) and later "# mount -a"
> 
> /dev/ocfs2/ocfs_01   /ocfs_01                ocfs2    _netdev,
> datavolume,nointr        0 0
> 
>  
> 
> PS: Create a lvm and format just on one node, at another nodes only
> include on fstab and enjoy
> 
>  
> 
> PS2: The attached document was written in Portuguese, so if you have
> any problem or doubt to understand, reply.
> 
>  
> 
> Any doubt or "append" for this procedure, reply :)
> 
>  
> 
> Good luck
> 
> Waldirio
> 
> 
> 
> 
> 2008/7/3, Dominik Vogt <dominik.vogt@xxxxxxxxxxxx>: 
>         Hello Waldirio,
>         
>         No doubt's, it's just a fact that I cant boot from ocfs2. My
>         systems are
>         all debian etch. So just for understanding what I did (Maybe I
>         understood all wrong about ocfs2)
>         
>         So I create a Volume Group, in this case it is about 1.5TB.
>         Then I
>         create vor every domu a Logical Volume, and format it with
>         ocfs2
>         filesystem. After that i install a system on that Logical
>         Volume, it's
>         working until this. but if I want to start the domu, xen
>         starts the
>         system but on boot it can't find the root partition! so it
>         seems like
>         the kernel can't initalize the ocfs2 fs during startup.
>         
>         Do I have to do some special things to get it running?
>         
>         Thanks and A nice day,
>         Dominik
>         
>         Am Mittwoch, den 02.07.2008, 14:01 -0300 schrieb Waldirio
>         ManhÃes
>         Pinheiro:
>         > Hello Dominik
>         >
>         > What you doubt about ocfs2 implementation (all doubt or at
>         specific
>         > point) ?!, and what the Operating System ?!
>         >
>         > Cheers
>         > Waldirio
>         >
>         > 2008/6/27 Dominik Vogt <dominik.vogt@xxxxxxxxxxxx>:
>         >         Dear List,
>         >
>         >         I'm thinking about using xen productive in our
>         datacenter, i'm
>         >         still
>         >         testing around with it. Now I got some questions,
>         just for
>         >         basic
>         >         understanding, we got for example this environment:
>         >
>         >         2 Nodes
>         >         1 SCSI Pool server (Connected via scsi to both
>         nodes)
>         >
>         >         Now I want to build a "cluster" so i would like to
>         make this:
>         >
>         >         Node 1 -> Primary   -|                          |
>         --> domU 1
>         >         (LV/OCFS2)
>         >                             |-> SCSI POOL 1.5TB (VG)   | -->
>         domU 2
>         >         (LV/OCFS2)
>         >         Node 2 -> Secondary -|                          |
>         --> domU 3
>         >         (LV/OCFS2)
>         >
>         >         So the reason, why I use ocfs2 is, because I access
>         the same
>         >         Disk with 2
>         >         Nodes via scsi. It would destroy the fs, if I use
>         ext3
>         >         (normaly).
>         >         but.... I can't boot the domU's form a ocfs2
>         filesystem... So
>         >         I tried to
>         >         setup a domU and installed ocfs2 into it, after that
>         I cloned
>         >         the system
>         >         into a ocfs2 lv, but if I start it, it can't find
>         the root
>         >         filesystem.
>         >         what else do I have to attend to get a system
>         startet from a
>         >         ocfs2
>         >         filesystem? Or is this the wrong way?
>         >
>         >         Regards, D. Vogt
>         >
>         >
>         >
>         >         _______________________________________________
>         >         Xen-users mailing list
>         >         Xen-users@xxxxxxxxxxxxxxxxxxx
>         >         http://lists.xensource.com/xen-users
>         >
>         >
>         >
>         > --
>         > ______________
>         > Atenciosamente
>         > Waldirio
>         > msn: wmp@xxxxxxxxxxxxx
>         > Site: www.waldirio.com.br
>         > Blog: blog.waldirio.com.br
>         > PGP: www.waldirio.com.br/public.html
>         --
>         Dominik Vogt
>         Informatikbetreuer
>         UniversitÃt Bern
>         Institut fÃr medizinische Lehre (IML)
>         Abteilung fÃr Assessment und Evaluation (AAE)
>         Inselspital 37a
>         CH-3010 Bern
>         
>         Phone: 031 632 3584
>         Mail: dominik.vogt@xxxxxxxxxxxx
>         
> 
> 
> 
> -- 
> ______________
> Atenciosamente
> Waldirio
> msn: wmp@xxxxxxxxxxxxx
> Site: www.waldirio.com.br
> Blog: blog.waldirio.com.br
> PGP: www.waldirio.com.br/public.html 
> _______________________________________________
> Xen-users mailing list
> Xen-users@xxxxxxxxxxxxxxxxxxx
> http://lists.xensource.com/xen-users
-- 
Dominik Vogt
Informatikbetreuer
UniversitÃt Bern
Institut fÃr medizinische Lehre (IML)
Abteilung fÃr Assessment und Evaluation (AAE)
Inselspital 37a
CH-3010 Bern

Phone: 031 632 3584
Mail: dominik.vogt@xxxxxxxxxxxx

Attachment: signature.asc
Description: Dies ist ein digital signierter Nachrichtenteil

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.