Pool size reporting is off in 9.2.1.2 rc/release?

[bigphil]

I just noticed when upgrading from 9.2.1.1 to 9.2.1.2-RC-6e25d6f-x64 and 9.2.1.2 Release, the Pool size that is being reported in the GUI is not correct now. In a mirrored pool with two 1.5 TB drives that used to report correctly, now report as 994.9 GB Pool size. Can anyone else reproduce this or am I missing something that I should look at? zpool status is showing healthy, smart is showing no errors, and there is no quota on the pool. The disk sizes are reporting correctly, just not the pool now.

 

[cyberjock]

Can you post the output of zpool list please?

 

[bigphil]

Code:

NAME    SIZE  ALLOC  FREE    CAP  DEDUP  HEALTH  ALTROOT
Pool1  1.36T  374G  1018G    26%  1.00x  ONLINE  /mnt

1.36T is what the GUI used to show on 9.2.1.1. Now it says 994.9.

 

[eraser]

Yup me too after upgrading to 9.2.1.2. GUI incorrectly shows my pool as 1 TiB in size. (see attached screenshot)

"zpool list" shows correct size:

[root@bfd] ~# zpool list
NAME SIZE ALLOC FREE CAP DEDUP HEALTH ALTROOT
tank 1.59T 541G 1.06T 33% 1.00x ONLINE /mnt

 

[bigphil]

Bug report created: https://bugs.freenas.org/issues/4419

 

[eraser]

cool thanks for opening the Bug. I have added my notes to it also.

 

[Serverbaboon]

Is this just a gui display bug that only matters if you want edit the pool or is causing issues elsewhere, was about to install 9.1.1.2.

 

[Schwinni]

Since "zpool list" shows the correct sizes it has to be a bug in the GUI only.

 

[GaiusBaltar]

interesting. i've got a 6x2TB RAIDZ2 pool. the GUI used to report 7.1TB total capacity for the entire pool. yet some datasets would report 6.8TB (this was even back in 9.2.0) now the pool shows 6.8TB in the GUI (one dataset still shows 7.1TB)
yet when i run "zpool list", i see the following:

Code:

[root@freeNAS] /# zpool list
NAME        SIZE  ALLOC  FREE    CAP  DEDUP  HEALTH  ALTROOT
ZFS_VOL00  10.9T  482G  10.4T    4%  1.00x  ONLINE  /mnt
 
[root@freeNAS]# df -h | grep ^ZFS_VOL00
ZFS_VOL00                                                      6.8T    431k    6.8T    0%    /mnt/ZFS_VOL00
ZFS_VOL00/.system                                              6.8T    335k    6.8T    0%    /mnt/ZFS_VOL00/.system
ZFS_VOL00/.system/cores                                        6.8T    26M    6.8T    0%    /mnt/ZFS_VOL00/.system/cores
ZFS_VOL00/.system/samba4                                      6.8T    4.3M    6.8T    0%    /mnt/ZFS_VOL00/.system/samba4
ZFS_VOL00/.system/syslog                                      6.8T    1.4M    6.8T    0%    /mnt/ZFS_VOL00/.system/syslog
ZFS_VOL00/PLEX                                                7.1T    273G    6.8T    4%    /mnt/ZFS_VOL00/PLEX
ZFS_VOL00/jails                                                6.8T    61M    6.8T    0%    /mnt/ZFS_VOL00/jails
ZFS_VOL00/jails/.warden-template-pluginjail-9.2-RELEASE-x64    6.8T    1.4G    6.8T    0%    /mnt/ZFS_VOL00/jails/.warden-template-pluginjail-9.2-RELEASE-x64
ZFS_VOL00/jails/plexmediaserver_1                              6.8T    10G    6.8T    0%    /mnt/ZFS_VOL00/jails/plexmediaserver_1
ZFS_VOL00/jails/sickbeard_1                                    6.8T    2.0G    6.8T    0%    /mnt/ZFS_VOL00/jails/sickbeard_1
ZFS_VOL00/jails/transmission_1                                6.8T    1.8G    6.8T    0%    /mnt/ZFS_VOL00/jails/transmission_1
ZFS_VOL00/user1                                                6.8T    12G    6.8T    0%    /mnt/ZFS_VOL00/user1
ZFS_VOL00/user2                                                6.9T    20G    6.8T    0%    /mnt/ZFS_VOL00/user2

pool is healthy, i just am failing to see why there's (always been) a 2.8TB difference between pool size in the GUI vs cmdline.

 

[cyberjock]

Dammit, I was tired.. I meant "zfs list".

Explanation is https://bugs.freenas.org/issues/4419#note-4

 

[Schwinni]

Anyway, wrong sizes in the GUI aren't new in 9.2.1.2.
I'm still on 9.2.0 and the GUI shows me:

Code:

Name      | Used      | Available  | Size
-----------------------------------------------
volume1  | 320.0 KiB | 5.9 TiB    | 5.9 TiB
dataxchg | 288.0 KiB | 1024.0 GiB | 1024.0 GiB
media    | 6.5 TiB  | 5.9 TiB    | 12.5 TiB

zfs list gives me:

Code:

NAME              USED  AVAIL  REFER  MOUNTPOINT                         
volume1          6.54T  5.91T  320K  /mnt/volume1                       
volume1/dataxchg  288K  1024G  288K  /mnt/volume1/dataxchg               
volume1/media    6.54T  5.91T  6.54T /mnt/volume1/media 

When summing up the values of USED and AVAIL the sizes are correct.

EDIT:
I think the GUI just sums up AVAIL and REFER. That's why the total size of the volume is incorrect. REFER of the volume doesn't reflect the values of REFER of the datasets.

 

[cyberjock]

Possibly. Datasets are like separate pools in some ways, but not in others. That's one reason why so many people get confused. There's a very long winded discussion on this topic somewhere via Google for those that want to understand the intricate details.

 

[GaiusBaltar]

Dammit, I was tired.. I meant "zfs list".

Explanation is https://bugs.freenas.org/issues/4419#note-4

you, good sir, deserve an internet high-five. thanks for the great explanation.

the output below now looks much more reasonable and understandable.

Code:

[root@freeNAS] /# zpool list
NAME        SIZE  ALLOC  FREE    CAP  DEDUP  HEALTH  ALTROOT
ZFS_VOL00  10.9T  482G  10.4T    4%  1.00x  ONLINE  /mnt
 
[root@freeNAS] /# zfs list
NAME                                                          USED  AVAIL  REFER  MOUNTPOINT
ZFS_VOL00                                                    321G  6.82T  432K  /mnt/ZFS_VOL00
ZFS_VOL00/.system                                            32.4M  6.82T  336K  /mnt/ZFS_VOL00/.system
ZFS_VOL00/.system/cores                                      26.6M  6.82T  26.6M  /mnt/ZFS_VOL00/.system/cores
ZFS_VOL00/.system/samba4                                    4.32M  6.82T  4.32M  /mnt/ZFS_VOL00/.system/samba4
ZFS_VOL00/.system/syslog                                    1.20M  6.82T  1.20M  /mnt/ZFS_VOL00/.system/syslog
ZFS_VOL00/PLEX                                                274G  6.82T  274G  /mnt/ZFS_VOL00/PLEX
ZFS_VOL00/jails                                              12.0G  6.82T  61.3M  /mnt/ZFS_VOL00/jails
ZFS_VOL00/jails/.warden-template-pluginjail-9.2-RELEASE-x64  1.34G  6.82T  1.34G  /mnt/ZFS_VOL00/jails/.warden-template-pluginjail-9.2-RELEASE-x64
ZFS_VOL00/jails/plexmediaserver_1                            9.40G  6.82T  10.7G  /mnt/ZFS_VOL00/jails/plexmediaserver_1
ZFS_VOL00/jails/sickbeard_1                                  759M  6.82T  2.03G  /mnt/ZFS_VOL00/jails/sickbeard_1
ZFS_VOL00/jails/transmission_1                                434M  6.82T  1.76G  /mnt/ZFS_VOL00/jails/transmission_1
ZFS_VOL00/user1                                              14.6G  6.82T  12.4G  /mnt/ZFS_VOL00/user1
ZFS_VOL00/user2                                              20.9G  6.82T  20.9G  /mnt/ZFS_VOL00/user2

thanks again.

 

[bigphil]

Something still seems amiss in the GUI to me. My Pool is a two disk mirror and the GUI reports the disks at 1.5 TB each. "zpool list" shows it correctly being 1.36TiB in size, and "zfs list" is only showing the pool available space, not the total size. The GUI is reporting the size and available space from zfs list, but in my opinion, they should be two different calculations (like they were in 9.2.1). I would think the size column would report on the size of the Pool from zpool list and any datasets/zvols size would be reported the same unless they have a quota and then report on that size. The available column should be calculated on zfs list though. The available column and size column shouldn't match. Whatever they had in 9.2.1 seemed correct to me (and at one point to the dev's too ;-).

 

[cyberjock]

Actually, your 1.5TB "shrunk" to 1.36TB because of the conversion from a base-10 "TB" to a base-2 "TB"(also referred to as TB to TiB).

Go to http://www.unitjuggler.com/convert-memory-from-TB-to-TiB.html and out in 1.5TB on the left. The number you'll get out is 1.3642420526594 TiB.

 

[bigphil]

Thanks for the clarification on that one...I wasn't even aware of such a unit of measure.

 

[ndboost]

just to chime in, im on 9.2.0 and i have this issue after doing a snapshot copy to a new pool. After further reading it seems its been dated back to v8~