I know the ReadyNAS products are "EOL", but such resources are still useful for those of us with units that are working absolutely fine.
I found this knowledge base article, which describes the principles of RAID, but not how varied disk capacities contribute to the final XRAID capacity.
Is there a similar calculator, or perhaps, Netger might consider restoring the old one just for reference?
I did find this thread (from 2011), which I assume remains valid to do it manually:
If you have single redundancy, you remove one of the largest drives. So in the case 4x1.5TB drives and 2x3TB drives, it is 4x1.5x.93=5.6TB and 1x3TBx.93=2.8TB which equals 8.4TB. If you were 3 and 3, it would be 3x1.5TBx.93=4.2TB plus 2x3TBx.93=5.6TB for a total of 9.8TB. In the case of Dual redundancy you must have 4 of the larger drives for the second layer to come in, so in the either 4x1.5TB and 2x3TB or 3 and 3 you would have 4x1.5TBx.93=5.6TB.
I found this knowledge base article, which describes the principles of RAID, but not how varied disk capacities contribute to the final XRAID capacity.
If you don't want to waste disk space, then with single redundancy you need to make sure the two largest disks are always the same size. For dual redundancy, you need to make sure the four largest disks are the same size.
If you follow these rules and start from a factory default, the rules for XRAID are easy to state:
Single redundancy - "sum the disks and subtract the largest"
Dual redundancy - "sum the disks and subtract the largest two"
The old X-RAID calculator never handled expansion, and often that created confusion when the volume size was smaller than expected after expansion..
If you want to avoid wasting space when expanding, then it is alway safe to
Add new disks that are at least as large as the largest already in the NAS
Single redundancy: make sure the largest two disks are the same size
Dual redundancy: make sure the largest four disks are the same size
If you know the details of your system's RAID groups, there are other sometimes other options - but it is best to address them case by case. The formulas would be unwieldy, and they would depend on the expansion history of your volume.
This is the correct capacity. Although if you are thinking that the 6 TB redundancy is all on one disk, then that is incorrect. The parity blocks are distributed across all the disks. If you put all the parity blocks on one disk, then the write speed of that disk becomes a performance bottleneck. Plus distributing the parity blocks evens out the load on the disks in the volume.
You do need to be careful with it. While it will give you the correct size for a new volume, it does not model expansion correctly. So it can give incorrect results sometimes when you are expanding to the configuration you enter.
But if you follow the basic rule to always replace a disk with one of the same size, or replace it it with one that is at least as large as the biggest disk in the array, then it works ok.
IMO it's quicker just to do the math of "sum the disks and subtract the largest".
