Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

For the last year I've been using a piece of software named unRAID built on top of slackware to host my NAS. I feel like I want to move on and evolve and build my own NAS setup on top of another distribution and have it really tailored towards me; issues are already rolling it.

First of all, I wish to have a software raid 4, with easy growth and difference sized disk. So far, all I've been able to find are people saying 'Just partition the disks so they're all the same size' which isn't very useful. The largest disk will obviously be the dedicated parity (Raid 4), and, I don't see why I can't have smaller disks run under it (Just assume any bytes > the disk size are zero for the calculation purposes).

Can anyone think of a way to do this? I care less about the read/write performance than the redundancy, so, even if it's some weird way of having a unique file system per disk, then, just mashing up the drives to create the parity (Which I believe is what unRAID does), then, I'm fine with that too (In fact, I'd prefer it).

share|improve this question
    
Is there any reason you want to use RAID 4? It is rather exotic. Most people with linux and software RAID seem to use mdadm to build a single RAID5 array, a pair of RAID1's. See this link for RAID level explanations. –  Hennes Sep 3 '13 at 10:52
    
Mainly because I just don't simply see how on earth I would have raid {anythingelse} with different sized drives, other than raid 0 which is not what I want. Here's the examples:- | Raid 1: If I have one drive that's 8GB, and one drive that's 4GB, the data from the 8GB won't be able to copy over to the 4GB, or, 4GB from the 8GB will be wasted. | Raid 2, 5, 6:- Invalid due to the fact they need the largest drive, the parity, spread over all disks, and thus, they need all disks the same size. | Raid 3&4:- Both have a dedicated parity and would work assuming the parity is the largest disk. –  Automatic Sep 3 '13 at 16:51
    
I might be wrong, I'm not 100% sure. I just see that if we have unique file systems for each disk (E.G. works for whatever size), then just calculate the parity size based on those disks (In real time, of course), then, any blocks which are in a larger index than the drives disk, assume as zero. This means if the drive ever fails and you install a larger disk, the front of the drive will be setup with the old filesystem with a bunch of plaintext at the end. Just increase the partition and grow the filesystem. Easy. –  Automatic Sep 3 '13 at 16:57
add comment

Know someone who can answer? Share a link to this question via email, Google+, Twitter, or Facebook.

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.