ZFS Is the zpool size an integer multiple of GB?

No.​
  1. Bash:
    truncate -s $((64 * 1024 * 1024)) testpool
  2. Bash:
    zpool create $_ `pwd`/$_
  3. Code:
    zpool get -p size testpool
    reports 50331648. That is 48 MiB, i. e. not even a whole GB or GiB.​
Why would you even care?​
 
Why use the -p flag if you want human-readable numbers? zpool-list(8) without any arguments shows all numbers with M/G/T[...] suffixes.

And to answer your question: as pretty much any other tool, zpool- p also uses bytes as the standard, 'non-suffixed'/'non-human' unit. This prevents any misconceptions about the 1000/1024 multiples (e.g kB/MB/GB vs kiB/MiB/GiB)
 
I use a 476g partition and a 475g partition to create a zpool.
They are the same size, 472g
29797.jpg
 
don't post pictures of text; its extremely inconvenient and wasteful. use the [code][/code] tags.

also: give the output of gpart show for each partition scheme and give the pools different names to avoid any stale cache/metadata issues (and/or issue a zpool labelclear after pool destruction)
 
don't post pictures of text; its extremely inconvenient and wasteful. use the [code][/code] tags.

also: give the output of gpart show for each partition scheme and give the pools different names to avoid any stale cache/metadata issues (and/or issue a zpool labelclear after pool destruction)
Sorry, typing manually is too trouble.
Code:
# zpool destroy r
# zpool labelclear nda1
# zpool create -m none rp nda1p1
zpool get -p size
NAME PRIPERTY VALUE SOURCE -
rp size 506806140928 -
29807.jpg

it's the same result
 
gpart show:
The output is the logical starting block, the partition size in blocks, the partition index number, partition type and human readable size.

Are you asking why zpool size property is not the same as the partition size?
Example from my single disk, single pool system. Note that gpart show has size of
484200448 blocks (gpart sized blocks) and human readable of 231G.
zpool list shows size of 230G and zpool get -p size shows 246960619520 bytes which converting to G is 230G.

root@FBSDNUC:~ # gpart show
=> 40 488397088 ada0 GPT (233G)
40 1024 1 freebsd-boot [bootme] (512K)
1064 984 - free - (492K)
2048 4194304 2 freebsd-swap (2.0G)
4196352 484200448 3 freebsd-zfs (231G)
488396800 328 - free - (164K)

root@FBSDNUC:~ # zpool list
NAME SIZE ALLOC FREE CKPOINT EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT
zroot 230G 65.3G 165G - - 51% 28% 1.00x ONLINE -

zpool get -p size
NAME PROPERTY VALUE SOURCE
zroot size 246960619520 -

Your zpool get -p size of 506806140928 divide out and you wind up with 472G.
A filesystem size (I'm using filesystem to also mean zpool) is never (almost never?) going to be the exact size of a partition because of the metadata needed to define the filesystem. Larger partition for larger filesystem means more metadata so more variance between filesystem useable space and raw partition size.

Please show the output of "zpool show" and compare the human readable size.
 
Back
Top