On 3/1/2019 6:04 PM, Kevin Locke wrote: > Presumably the error occurs because parted is using the optimal_io_size > for alignment. Since optimal_io_size is based on USB constraints rather > than disk constraints, it does not seem suitable for partition layout. > > I would suggest ignoring optimal_io_size if it is not a multiple of > minimum_io_size, as done by cryptsetup[1]. Alternatively, 33553920 > could be ignored specifically, since it seems common for USB disks. > > This issue has been discussed elsewhere: > https://www.saout.de/pipermail/dm-crypt/2016-January/004934.html > https://bugzilla.redhat.com/show_bug.cgi?id=1513820 > https://unix.stackexchange.com/q/340484 > https://linux-blog.anracom.com/2018/12/03/linux-ssd-partition-alignment-problems-with-external-usb-to-sata-controllers-i/
Interesting. It sounds like USB and MD are using optimal_io_size in fundamentally incompatible ways. The question then is: is USB doing the wrong thing ( and should be fixed ), or is the very definition of optimal_io_size fundamentally broken?