Sign in to follow this  
Followers 0
Daleynoy

Raid overhead and best practices

3 posts in this topic

Hello everyone

I have a question that is fairly wide open and I imagine the exact answer will be "well, it depends" This is really a getting started question and I expect I'll be going in multiple directions for answers. So here we go.

Is there a recommended free space percentage when setting up raids. Any by that I mean, I have a six 300GB drives, I configure them for for a raid 5. When I create the logical volume, is there a reason to not allocate all 1.5TB. Say I only allocate 80% of that space, so let's say I allocate 1.3TB and leave 200GB of space out there. Does the raid controller\System ever use that 200GB of space for recovery blocks etc etc.

Now I know there are probably a dozen questions people have on that. Is it hardware or software, who's the vendor, Is it SCSI\ATA\SAS disks. is it virtual arrays. I am looking at more tradtional style raids of hardware and software and a little less on the new Virtual arrays, but I'm interested in it all.

The answer to those questions are, you tell me. I'm looking for any example of where a software SATA raid would want\need to have 20% unallocate space but a SCSI hardware raid does not or maybe a Software SCSI does, but a Hardware SATA doesn't.

Again a fairly wide open question and I'm just looking to understand the different needs of the different style of raids. Any pointers\links that you have from various vendors would be wonderful.

Thanks

Edited by Daleynoy

Share this post


Link to post
Share on other sites

No. Overprovisioning only serves a purpose on SSDs. Any space that you don't define as a part of a logical device on a platter RAID controller would simply be wasted. The RAID controller will automatically take whatever metadata it needs out of the available total. There is no difference between SATA and SAS in this context.

Now you can increase the recovery speed and fault-tolerance by using RAID 6 instead of RAID 5. You will get one drive's worth less available space, but security against two-drive failures and faster rebuilds of failed drives.

Share this post


Link to post
Share on other sites

Now you can increase the recovery speed and fault-tolerance by using RAID 6 instead of RAID 5. You will get one drive's worth less available space, but security against two-drive failures and faster rebuilds of failed drives.

I am still very new to RAID but in my opinion you are mistaken, RAID6 uses dual parity calculation to provide resistance against two HDD failures.

Dual parity requires the controller to do twice the calculations which increases rebuild times.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0