isopropyl
Contributor
- Joined
- Jan 29, 2022
- Messages
- 159
So, your pool's total space is now 25.4 TB. usable space is 20.3 TB. If we add a single n-way mirror vdev (it's not important how wide the mirror is, the space increase is the same) made of 20 TB drives, we have to add 20 TB * 0.8 = 16 TB of usable space.
So that means 20.3 + 16 = 36.3 TB of usable space with ALL your 7 vdevs of 4TB drives + the new mirror of 20TB drives; the spares do never count because they don't add any space.
Now, of that 36.3 TB you would be using 19.3 TB for your current data: 36.3 - 19.3 = 17 TB of free, usable space.
Each 4 TB mirror vdev adds to the pool 4 * 0.8 = 3.2 TB of usable space, as such: 4 (vdevs) * 3.2 = 12.8 TB
If we remove those 12.8 TB from the 17 TB of free, usable space we get 4.2 TB of free, usable space with the single n-wide mirror vdev of 20 TB drives and 3 2-way mirror vdevs of 4TB drives, plus the couple of spares.
It took me a couple reads, but I think I understand. I kept losing you at "Now, of that 36.3 TB you would be using 19.3 TB for your current data: 36.3 - 19.3 = 17 TB of free, usable space.", however I see why we went further than that. I didn't realize you were calculating all my current drives plus the new 20tb drives. So you calculated it as a whole, then after you removed the 4 vdevs and got 12.8TB. Makes sense.
So simply put, after adding the 20tb drives, and removing the 4 devs, I will be left with 4.2tb of usable space.
However, currently I have 5.9TB of UNUSED free space in my pool. So I can safely add on 4.2TB to that correct? And I would have 10.1TB of UNUSED free space, correct?
By unused, I mean no data on it. Free. I can put stuff on it. Keyword, unused.
Or would that 5.9TB actually lower itself to 4.2TB, because of how you did the math?
P.S.: both of your spares are currently in use inside different vdevs.
Are 2 of My Drives Failed? (See Edit: Moving Data Onto To New Vdev, To Remove Old)
Normally I wouldn't suggest weekly scrubs, but we are outside of standard procedures. You could try to reach to HGST (or WD?) in order to see if they have a way to fix this but I wouldn't be very optimistic. I suggest you to update the firmware of your HBA anyway. Edit: found this. It has a...

I never really figured it out, I was waiting for two replacement drives to come in the mail just in case.
This does scare me a little though, because they are in use how would I go about choosing which hard drives to remove?
Worst case scenario, I take them out, and it copies the data but if something fucks up, I just put the 4tb drives back in no issues, no?
When "lose a drive" is stated when it comes to mirrors or raid, that implies the drive actually dying/dies correct? If I simply unmounted/disconnected 2 healthy drives in an entire vdev, nothing bad would actually happen, the pool would just be dead, but when I plug them back in, the pool would return to normal, correct?
Like say I disconnect both of my two drives in mirror 1.
Nothing bad actually happens correct? As long as the data on the two drives stays in tact, I can simply plug that mirror back in and all is good.