Hi, I'd like to ask for some advice, as I'm mulling amongst 2 different options.
I have a 7-disk raidz2 pool, using 8TB 7200RPM HDD that is at 97% capacity (I know I know, should keep below 80% for performance, but it's mostly video, so don't really need a lot of raw performance, price:capacity ratio is a bigger factor than raw performance).
I have 5 unused 8TB 7200RPM HDD available.
What would be the better way to go, create a 2nd 5-disk raidz2 pool, and migrate some of the (still growing) datasets to it from the first pool, or buy 2 more HDD so I'd have 7 disks to add to the existing pool as a second 7-disk raidz2 VDEV? I understand that the best practice with multiple VDEVs is for them to be identical in terms of number of disks, vdev type, capacity, and performance, hence buying 2 more disks to match those characteristics with the existing VDEV.
The latter option would save a lot of work in doing dataset migrations between pools, and easier to do other fancy stuff if I decide to do it later like adding a metadata special vdev or other similar configurations since I'd only have one pool for that to apply to, rather than multiple pools each requiring their own independent configurations and devices. But it has obvious downside of immediately costing money I otherwise wouldn't spend in buying an extra 2 HDDs (although the money isn't a problem if that's the better way to go).
Are there any other options I'm missing? Thoughts on which would be the better option?
_____________________________________________________________________
Supermicro X11SSL-CF
Intel E3-1240L v5
32GB ECC RAM
Intel RES3TV360 expander
Linux 5.15.52-gentoo
I have a 7-disk raidz2 pool, using 8TB 7200RPM HDD that is at 97% capacity (I know I know, should keep below 80% for performance, but it's mostly video, so don't really need a lot of raw performance, price:capacity ratio is a bigger factor than raw performance).
I have 5 unused 8TB 7200RPM HDD available.
What would be the better way to go, create a 2nd 5-disk raidz2 pool, and migrate some of the (still growing) datasets to it from the first pool, or buy 2 more HDD so I'd have 7 disks to add to the existing pool as a second 7-disk raidz2 VDEV? I understand that the best practice with multiple VDEVs is for them to be identical in terms of number of disks, vdev type, capacity, and performance, hence buying 2 more disks to match those characteristics with the existing VDEV.
The latter option would save a lot of work in doing dataset migrations between pools, and easier to do other fancy stuff if I decide to do it later like adding a metadata special vdev or other similar configurations since I'd only have one pool for that to apply to, rather than multiple pools each requiring their own independent configurations and devices. But it has obvious downside of immediately costing money I otherwise wouldn't spend in buying an extra 2 HDDs (although the money isn't a problem if that's the better way to go).
Are there any other options I'm missing? Thoughts on which would be the better option?
_____________________________________________________________________
Supermicro X11SSL-CF
Intel E3-1240L v5
32GB ECC RAM
Intel RES3TV360 expander
Linux 5.15.52-gentoo