Question about HDDs before I start using UNRAID


Recommended Posts

Hey Guys, I have a question about Unraid. On my storage Server I currently run Linux with ZFS. In the Server there is currently installed:

1 NVME Drive for the OS

3x SSD for "faster Storage" in raidz-1

 

4x 3TB HDDs in raidz-1

2x 4TB HDDs als mirror

Both are currently combined as a stripe

 

One of the 3TB drives is failing now.

I have one 10TB drive currently here and two 4TB drives are ordered and going to arrive in a few days.

 

My Plan is to thow out the faulty 3TB drive and install the 10TB drive. First of all, I'm going to copy the data to the 10TB drive (I have backups on tape, so don't worry) and then install Unraid on the machine, create a new raid out of the 3x3TB and 2x4TB drives and copy the data back. And when the 2x 4TB drives arrive, I'm going to add them to the "pool". 

 

Is this possible? Or do I have to rebuild the pool and copy the data again after I added the drives?

Link to comment

Yeah, I also noticed this problem after a little bit of research ;)

But I also have two 5TB drives in another machine, so I can use them within the array.

 

But there is a thing, I don't understand. Two things are important for me:

 

First of all is the performance. So, I have the 1 NVME drive and 3 SSDs in the machine. I can use them as cache. But how is the performance going to be with the other HDDs? What kind of reading speed can I get when I don't use the cache?

 

The other important thing is the encryption of the data. On my linux machine I have the data Pool mounted to /data1 and everything that goes into /data1/enc is going to be encrypted. 

 

On Unraid I have to use encrypted XFS and also encrypt the Cache and the Parity FS as well. So it's going to slow the performance again a little bit, am I right?

Link to comment

Parity doesn't have a filesystem so encryption doesn't apply to parity.

 

Unraid IS NOT RAID isn't necessarily a problem, it is just another way of doing things, which has tradeoffs.

 

There is no striping in Unraid. Each data disk in the parity array is an independent filesystem which can be read all by itself on any Linux. Folders can span disks (user shares), but each file is completely contained on a single disk. Read speed for any file is at the speed of the single disk containing the file.

 

Unraid allows 1 or 2 parity disks, which provide redundancy to recover 1 or 2 simultaneous disk failures. Parity is realtime, so writes to the parity array are somewhat slower than single disk speed since parity also has to be updated. There are 2 different methods for updating parity, each has its own tradeoffs.

 

And, since each disk in the array is independent, if for some reason you lose more than parity can recover, you will still have whatever disks can still be read. Because each data disk in the parity array is independent, you can mix different sized disks in the array, and easily replace or add additional disks without rebuilding the whole array.

 

Unraid also provides for faster storage in the cache pool, and various btrfs raid configurations are supported. These are usually SSDs where files that need performance will be stored (dockers and VMs), and where writes to the user shares will be temporarily saved until moved to the slower parity array.

 

  • Like 1
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.