Mixing drive makes/models within a NAS

If you have mixes disk sizes how does the redundancy work?

If the disks are all sames I understand how it works, but if you have 3---2tb and 1---4tb then how can redundancy be spread equally across the 4 disks?

RAID 5. This RAID level also provides data redundancy, but it requires at least three disks. RAID 5 uses

the capacity of one disk to protect you from data loss if one disk fails.Your data is distributed across

multiple disks to improve disk performance. The total capacity of your storage system equals the capacity

of all your disks minus the capacity of one disk. It is supported on systems with at least four drive bays.

X-RAID uses the capacity of one disk for data storage and reserves the capacity of a second disk for data

protection, which allows the volume to recreate data if a disk fails. In a two-disk system, the usable storage

space is one disk. In a three-disk system, the usable storage space is two disks. In general, the total capacity

of your storage system equals the capacity of all your disks minus the capacity of one disk.

---

FG wrote:

If you have mixes disk sizes how does the redundancy work?

 

 

---

That's the reply you didn't understand.

You can think of it as a layer cake.  For example, imagine you have a 6 disk array, with 1 TB + 4 TB + 6 TB + 6 TB + 4 TB + 1 TB.  You can visualize this as

        XXX XXX          top 2 TB layer (drives 3,4)

    YYY YYY YYY YYY      middle 3 TB layer (drives 2-5)

ZZZ ZZZ ZZZ ZZZ ZZZ ZZZ  bottom 1 TB layer (drives 1-6)

 1   2   3   4   4   6  (disk drive number)

The XXX, YYY and ZZZs are all partitions, laid out as shown in the figure above.  ZZZ is on all the disks; YYY and ZZZ on the 4 TB disks; XXX,YYY, and ZZZ are on the 6 TB disks.

RAID protection is applied to each layer separately as shown below. This creates three RAID partitions A, B, and C - think of them as virtual disks - with 2 TB (A), 9 TB (B) , and 5 TB (C) of available storage.  Note the top layer A is RAID-1, because there are only two partitions in it.

A 2x2 TB RAID-1  2 TB storage + 2 TB mirror  

B 4x3 TB RAID-5  9 TB storage + 3 TB parity

C 6x1 TB RAID-5  5 TB storage + 1 TB parity

          Total 16 TB storage + 6 TB redundancy overhead

These three RAID partitions are then assembled into one volume.  All this is done with standard linux tools.

Thank you for explanation.

The newly added 4tb drive to bay 4 finished syncing.

The total storage shown is about 5.3tb

                3.06 from data

                1.52 from  snapshots

                888 gb from free space

The total disk capacity is about 9tb

Does that seem correct?

https://drive.google.com/file/d/0B2b_iVvNvbfJeVlvV2Y5M3FLUnM/view?usp=sharing

---

FG wrote:

 

The newly added 4tb drive to bay 4 finished syncing.

 

The total storage shown is about 5.3tb

                3.06 from data

                1.52 from  snapshots

                888 gb from free space

 

The total disk capacity is about 9tb

 

Does that seem correct?

 

---

The rule for capacity is "sum the disks and subtract the largest".  So in your case that's 6 TB (same as was before you upgraded the first drive).

Note that the ReadyNAS is reporting space in TiB, not TB.  Your total is ~5.44 TiB  -3.06 + 1.52 + (888.41/1024).  Converting that back to TB gives you 6 TB.

After you do the next disk, you will see an increase of 2 TB (to ~7.3 TiB).

Going back to the layer cake model - your top layer of 2 TB only has one disk in it.  So there's no way to create redundancy in that layer.  The NAS just doesn't use that space.  Upgrading the second disk lets it use RAID-1 for it.

Thank you Stephen for you explanation, it was very helpful.

This morning, 5am, I pulled the smaller 2tb disk from bay 3 of my NAS and replaced it with a 4tb Seagate IronWolf drive.

As soon as I did that, I got a message about  volume in a degraded state and Resyncing started.

All of that is what I was expecting.

Then about 30 minutes ago, 7 hours after I replaced the drive, I check on the GUI and it said about 20 minutes left to resync.

Shortly after that I get 2 message, 

• Volume volume1 health changed from Degraded to Redundant.

• Disk in channel 3 (Internal) changed state from RESYNC to ONLINE.

All of that is what I was expecting as well.

Now I check the GUI and it is showing another recync progress, I'm 53% with nearly 4hours to go.

All the drives on the GUI show and green light and when I hoover over each drive it says they are online.

Should I be worried, is this normal behavior?

This is the second disk upgrade correct?  So first it resyncs the existing RAID layer, and then it starts building the upper RAID layer.

So this is normal.

Yes, the 2nd disk to be upgraded.

OK!. Good to hear that this is normal. I thought I had some hardware that was dying.

Can I drop in the 3rd disk to upgrade or should I wait for this current sync to complete?

I'm guessing I should wait.

Definitely wait until the resync is done (and your volume expands).  Pulling out a disk during a resync generally destroys your data.

Hummmm....

But as soon as the 1st sync was done the volume expanded and I gained another 2tb of capacity/storage.

I will wait to be safe.