On 27/03/18 15:30, Gandalf Corvotempesta wrote: > 2018-03-27 14:49 GMT+02:00 Wols Lists <antlists@xxxxxxxxxxxxxxx>: >> Which you can't do just yet ... the new array will be the exact same >> size as the old one, with a large chunk of unused disk space. LVM will >> see that the array hasn't grown and will barf if you try and increase >> the file-system size. >> >> mdadm --grow /dev/mdX --size=max > > So, I have to create a new partitionin schema on newer disks (using > all available space), > then mdadm will replace the old disks (with the old partition) to the > newer one (with the newer partion) If you use my scheme - yes. > > After that, md0 is still "small" even if placed on a huge partition. > With --grow I can grow the array to use the new space available. > Yup. mdadm can only use the space of the smallest partition, and it won't realise that has changed unless you explicitly tell it (it shouldn't just assume, because for whatever weird reason you might add a bunch of large drives and then want to add another small one ...) > Then, a "pvresize" followed by "vgextend" and "lvextend" should do the last job. > I don't know LVM, but that sounds about right. Actually, if you're using LVM already, John Stoffel's idea is probably best, and seeing as you've got three spare slots to add the drives, Andreas' idea is also a good one. You pays your money and you takes your choice :-) Cheers, Wol -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
