[ale] can this install be saved?

leam hall leamhall at gmail.com
Wed Mar 4 11:26:12 EST 2015


That means your OS disk is a Single Point of Failure. If the OS is
under LVM can you add it in with software RAID? Then use the other 2
disks as HW or SW RAID.

Personally, I'm more the "start over as RAID 5 with all 4 disks" type.
But that's personal preference.

On Wed, Mar 4, 2015 at 11:16 AM, Todor Fassl <fassl.tod at gmail.com> wrote:
> A researcher turned in a couple of old Sun Fire X4150 machines. I am tasked
> with turning them into a sandbox for students to play around with slurm,
> condor, etc.   I asked a co-worker to install ubuntu server on them but at
> the time, I didn't know each machine had 4 147G disks. I believe she did a
> standard ubuntu install and just put it on the first disk.
>
> root at cartan:~# fdisk -l | grep G
> Disk /dev/sda: 146.7 GB, 146685296640 bytes
> Disk /dev/sdb: 146.7 GB, 146685296640 bytes
> Disk /dev/sdc: 146.7 GB, 146685296640 bytes
> Disk /dev/sdd: 146.7 GB, 146685296640 bytes
> Disk /dev/mapper/cartan--vg-root: 133.5 GB, 133513084928 bytes
> Disk /dev/mapper/cartan--vg-swap_1: 12.9 GB, 12880707584 bytes
>
>
> Should I start over and configure RAID or should a use the 3 empty disks to
> configure a DNFS? I have to configure a distributed network file system
> anyway so that when a student logs into one machine, he gets the same home
> dir as if he logged onto the other. I usually set up gluster and then make
> the gluster volume their home directory.
>
> So I could just use the 3 empty disks on each machine as a gluster home
> directory space and then do triple replication.
>
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo



-- 
Mind on a Mission


More information about the Ale mailing list