*

[jgreco]

Awesome random blogs. He's trying to test the difference between striped mirrors and raidZx by writing to it using /dev/random as a source? That's hilarious.

It is also puzzling because the process of resilvering starts at the root and then performs a traversal of the allocated blocks based on the metadata. This has the advantage of only analyzing used and allocated blocks, and it has to work that way because the checksum for a block is stored in the metadata block "above" it. You can do certain things to sabotage this process, of course.

Additionally, having more vdevs also means that I/O is being spread out amonst all of the vdevs, so on a per-vdev basis, if the workload is the same, more vdevs would make resilvering go faster.

Or, put differently, you shouldn't build a system that doesn't have a lot of extra oomph. Building a system that can just barely handle the expected peak workload is a fail.