Am 10.10.22 um 17:23 schrieb Pascal Hambourg:
On 10/10/2022 at 15:49, Reindl Harald wrote :
i am at creating new RAID1 stoarges on twice sized disks to replace
existing 4 drive RAID10
looking with fdisk and calculate twice din't end well and finally "dd"
the FS in the array stopped with around 2 MB too small
Which is probably the size of RAID metadata added at the beginning or
end of the partitions.
which should have been on the old RAID10 too :-)
anways, two pairs of identical machines and after one is done partition
table can be cloned
is the "30716928" MiB or MB and what takes fdisk with "+30716928M"?
MiB for both.
Array Size : 30716928 (29.29 GiB 31.45 GB)
Used Dev Size : 15358464 (14.65 GiB 15.73 GB)
so i should be good with +30716930M
2 MB extra size - i have 3 MB unpartitioned space on the end of the 2TB
disks and will make boot 400MB instead 500MB to get some headroom :-)
did i say that i hate it that M isn't strictly 1024 when it comes to IT?
"M" is never 1024
in software outputs it often still is
"M" (mega) was defined and should always be used as 1000000. If you mean
1048576, use "Mi" (mebi)
well, on the other hand one could expect to store 1TB data on a 1TB
drive.... with 100 GB drives the doifferences wasn't that large but
these days....
one of my storage servers has a 6.8T ext4 FS on a 4x4TB RAID10 and 8
versus 6.8..... :-)