6 drives, 2x3tb 4x2tb, 3tb show as 2tb WTF

What ReadyNAS model do you have, and what is the firmware version?

Did you try restarting? Note that the expansion from 2-3 TB will happen at the end of the process, it is the last step.

Also, if you just purchased you have 90 days of free phone support.

Also are you using X-RAID2 single-redundancy or dual-redundancy?

Are there any practical negatives to installing mixed size disks like this into the Pro 6?

kevikev wrote:
Are there any practical negatives to installing mixed size disks like this into the Pro 6?

If you have an existing array with empty slots, then you can only fill the empty slots with the largest installed size (or larger). Similarly, replacement disks either match their existing size, or are the largest size (or larger).

Dual Redundancy requires 4 drives of the largest size.

Forensic recovery might be harder, as the volume is built up of multiple RAID layers.

I've run a mix of 1.5 TB and 3 TB in my Pro 6 for a while now, with no issues. Currently it is 2x1.5 TB + 4x3TB.

Note for your setup your 3 TB will be useless until you get another 3 TB drive because of the dual redundancy (XRAID6) you need 3 disks of each drive size to live up to the statement "dual redundancy"
In your current setup you have 3,3,2,2,2,2. If you lose a 2 TB disk your okay another, 2 TB drive will cover it. If you lose a 3 TB drive disk nothing can cover it and live up to the statement that you have dual redundancy in the case that you lose any disk. Get another disk of 3 TB in size or replace (vertical expand) one of your 2 TB drive with another 3 TB drive. Im not sure if you have another slot for an extra disk so verticle expansion is possibly all you have left to deal with. There is nothing wrong with your NAS its reacting the way it was designed to react to minimize damage to data and to increase reliability.

Here look at the results from here:

http://www.kossboss.com/volsizecalcsite

Single Disk Redundancy
Disk Space with Raid 5 [base 10]: 11000.000 GB = 11.000 TB
Final Useable/Filesystem Size with Raid 5 [base 2]: 10172.736 GB = 10.093 TB

Dual Disk Redundancy
Disk Space with Raid 6 [base 10]: 8000.000 GB = 8.000 TB
Final Useable/Filesystem Size with Raid 6 [base 2] : 7391.003 GB = 7.333 TB

*** You should have 7.333 TB about ***

kossboss wrote:
you need 3 disks of each drive size to live up to the statement "dual redundancy"

Actually you need 4 disks as Stephen mentioned above. A RAID-6 layer requires a minimum of four disks.

Negative I believe that is not correct mdmg - in the over all process you would be correct, but in expansion that is incorrect. To get RAID 6 you need at least 4 drives, but to get mdadm to let you get another md after your md2 with RAID 6 you only need 3 disks.

You see Im talking about the extra MD that was created with the oversized disks (md3). You can create a RAID 6 out of 3 block devices....Here is my proof... how many disks are needed for RAID5? 3 right... well then why does our NAS expand the md3 after 2 disks. Look at logs for a setup of 2,2,2,3,3 with RAID 5 you will see that the md3 is expanded with only 2 block devices (the last two 1 TB big portions [2tb to 3tb] of the 3 TB drives are used). Here is a visual of it on readynas.com : http://www.readynas.com/?cat=53 Note that after the 2nd disk upgrade the Volume gets upgraded... Meaning that md3 expanded and is adding its space to the LV /dev/c/c.

However in this case I dont know if the 3 tb drives were put in first or last. If they were put in first he needs to back up all his data to remote place and then factory default the nas with the 3,3,2,2,2 setup and then upgrade one of the disks to 3tb.
If however the 3TB drives were put in last then the user just needs to add another disk to experience volume growth.

The best resolution for the issue is for the user to include the logs - please provide us a link to your zipped up logs (System-Status-Logs-Download All Logs). The user can put them in Dropbox public folder or any internet file sharing service and just provide us the link.

In other words - its hard to put this into words people can visual

To start raid 5 you need 3 disks but to expand raid 5 you only need 2 - with mdadm software
To start raid 6 you need 4 disks but to expand raid 6 you only need 3 - with mdadm software

kossboss wrote:

You see Im talking about the extra MD that was created with the oversized disks (md3). You can create a RAID 6 out of 3 block devices....Here is my proof... how many disks are needed for RAID5? 3 right... well then why does our NAS expand the md3 after 2 disks. Look at logs for a setup of 2,2,2,3,3 with RAID 5 you will see that the md3 is expanded with only 2 block devices (the last two 1 TB big portions [2tb to 3tb] of the 3 TB drives are used). Here is a visual of it on readynas.com : http://www.readynas.com/?cat=53 Note that after the 2nd disk upgrade the Volume gets upgraded... Meaning that md3 expanded and is adding its space to the LV /dev/c/c.

That animation is for X-RAID2 single-redundancy.

kossboss wrote:
In other words - its hard to put this into words people can visual To start raid 5 you need 3 disks but to expand raid 5 you only need 2 - with mdadm software To start raid 6 you need 4 disks but to expand raid 6 you only need 3 - with mdadm software

The thing is though in the first case you have single-redundancy. Say a RAID-5 layer of 6x2TB + RAID-1 (that could be converted to RAID-5 when more disks are added) of 2x1TB.

In the second case whilst you could do e.g. RAID-6 layer of 6x2TB + RAID-5 of 3x1TB that wouldn't be dual-redundancy any more, so four disks are required for the volume to expand (regardless of when new md devices are created). When you add a third 3TB disk it might create a degraded RAID-6 layer but in order to maintain dual-redundancy it wouldn't be able to make use of it to expand the volume till a fourth disk is added.

I recently expanded my Ultra 6 volume. I had 6x1.5TB disks installed and I'm using X-RAID2 dual-redundancy. I had to install 4x3TB drives before expansion took place.

kossboss wrote:
...Note for your setup...

2x3tb + 4x2tb was zeeboid's setup, not kevikev's. Kevikev was asking generally about drawbacks of mixed size arrays, he didn't specify any specific config.

In principle you can get dual redundancy with 3 drives (simply by extending RAID-1 to a triple mirror). It would not be normal RAID-6, but it would be dual redundancy. However, my understanding is similar to mdgm's. While it may be theoretically possible, XRAID-2 won't do it.