(Best) way to mount 3 harddisks to one mountpoint?

I’ll be purchasing 3 harddisks soon and I was wondering if it’s possible to set them all to the same mountpoint, for example “home/user/somemp”.
Where “home/user/somemp” is 3TB if the disks are 1TB each.

I’m asking cause the situation is like this, I got a fileserver on which I’m archiving a ton of files and I’d like to be able to access all archived files from the same mountpoint, if one disk dies on me I don’t mind losing the data on just that disk. But I don’t want to run the risk of losing all data if only one disk dies.

So is this possible? If so… how I do it?

The best way to do this is using raid not used it myself but there’s lot’s of info on the net if you google

Geoff

On 2008-10-20, Axeia <Axeia@no-mx.forums.opensuse.org> wrote:

> I’ll be purchasing 3 harddisks soon and I was wondering if it’s possible
> to set them all to the same mountpoint, for example “home/user/somemp”.
>
> Where “home/user/somemp” is 3TB if the disks are 1TB each.
>
> I’m asking cause the situation is like this, I got a fileserver on
> which I’m archiving a ton of files and I’d like to be able to access all
> archived files from the same mountpoint, if one disk dies on me I don’t
> mind losing the data on just that disk. But I don’t want to run the risk
> of losing all data if only one disk dies.
>
> So is this possible? If so… how I do it?

Look into LVM.


Elevators smell different to midgets

RAID 0 would give me some massive dataloss if some disk dies, RAID 1 just eats away a lot of space. RAID 5 I was considering… but I’ll have to look into how much CPU power this eats away (will have to be software based).

This will get the job done, and from what I can tell it should be easy to add even more room/harddisks to it. However I can’t find what happens if one of the disks dies… how much data is gone?
Or more importantly… which data is gone, if it’s directories + content it’s not too bad… but if some random files disappear out of some random folders it would be horrible to find out what’s missing

My apologies if the post above did not make any sense… I shouldn’t be allowed to post after midnight

LWM would not fulfill the resilience requirement. As far as the OS is concerned, the blocks of the disks are concatenated into one large “disk”. So your data could be anywhere on the three disks, even split across them in pathological cases. You can make the data migrate off a member disk, but that has to be explicitly commanded, and not possible in the case of a breakdown.

RAID5 should be fine. CPU these days have lots of spare grunt.

Thank you ken_yap for clearing a few things up, I’ll be going with a software based raid 5. I’m aware this would push back the total capacity by one disk, but the resilience is also what I’m after (this is a bit more than I need even).

Bleh I hate replying to myself… but it seems you can’t edit your own post here if it’s been more than 10 minutes since you’ve posted it.

I just wanted to give additional information for those who need to make a similiar choice. Below is what I wanted to add :stuck_out_tongue:

I’ve read elsewhere that someone with a mere 500MHz processor was pretty much guaranteed a software raid 5 would run just fine. As my server has an Athlon 64 3000+ it shouldn’t impose any problem, even though it’s also has other tasks (printing, torrents, webserver)