Will this way of increasing the amount of storage on ZFS work?

Question

ZFS

Good day!
I plan to get a file cleaner on 3 disks connected to a RAID array and I want to use ZFS (RAIDZ1) for this. But, for now, only 1TB drives are available, which will provide me with 2TB of storage in total.
Hence the question! Taking into account the principles of operation of such an array (ZFS), if in the future it occurs to me to increase the amount of storage, say up to 8TB (having bought 3 disks of 4TB each), can I do this by successively ( physically ) replacing the disks in the pool and save while having the data?
That is:
Step 1 - replace the first disk (1TB + 1TB +4TB) - the data should be preserved, but the total volume will remain 2TB.
Step 2 - we replace the second disk (1TB + 4TB +4TB) - the same, the volume remains the same (2TB).
Step 3 - replace the third drive (4TB + 4TB + 4TB) - and... what happens? Will the total volume grow to 8TB or stay at 2TB. And won't it all fall apart?
In theory, it seems, everything should be smooth, but how it will actually be is a question for me!
The option of creating another pool nearby (8TB) and further transferring data to it will not work - there will be nowhere to put the raid controller. Yes, and other ways to connect three screws in a mini-ITX mother do not come to mind.
In general, if someone did something similar, or how to solve such a problem in a human way - please share your experience! I will be grateful!
PS System Ubuntu Server 16.04 LTS

Answer 1

Try to simulate your situation in virtualbox. Theoretically, what you want to do should work. True, after attaching the last 4TB disk, nothing will happen by itself. You will need to do an export and import of the pool. Your situation is well demonstrated here: wiki.opennet.ru/%D0%A3%D0%B2%D0%B5%D0%BB%D0%B8%D1%...