i just created a second raid1 swap device, with partitions identical in
size to the first, all are aligned on 8sector boundaries, however mdstat
reports the size of the second as 1024bytes smaller.  offhand my guess is
raid1 is only using the portion of the underlying partitions that line up
with some larger blocksize.  i wonder if it doesn't even matter whether
i've created the partitions on 8sector boundaries, i wonder if raid1 just
optimizes for that anyway.  if i knew that to be true, i'd fuss less over
how they're created.  anyone got the inside scoop?  got a reference?

details:
>Disk /dev/sda: 3907029168s
>Sector size (logical/physical): 512B/512B
>Partition Table: msdos
>Disk Flags:
>Number  Start        End          Size         Type      File system  Flags
> 1      2048s        20989951s    20987904s    primary                raid
> 5      2539062512s  2560050415s  20987904s    logical
>Disk /dev/sdb: 3907029168s
>Sector size (logical/physical): 512B/512B
>Partition Table: msdos
>Disk Flags:
>Number  Start        End          Size         Type      File system  Flags
> 1      2048s        20989951s    20987904s    primary                raid
> 5      2539062512s  2560050415s  20987904s    logical

mdstat however reports their sizes thus:
>Personalities : [raid1]
>md5 : active raid1 sdb5[1] sda5[0]
>      10484736 blocks super 1.2 [2/2] [UU]
>md127 : active raid1 sdb1[1] sda1[0]
>      10485760 blocks super 1.2 [2/2] [UU]

and swapon -s reports their sizes as 4bytes less than mdstat:
>Filename                                Type            Size    Used
Priority
>/dev/md127                              partition       10485756
27968   -1
>/dev/md5                                partition       10484732
0       -2
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20190110/269d9d1e/attachment.html>