Upgrade ZFS pool to larger drives. Increase available capacity.

[kamiller42]

Had mirror ZFS pool with 2 4T drives. Upgraded to 2 8T drives. Created a new pool on new drives using all available space. Performed a zfs send/recv to clone the pool to new drives. Content intact but pool & datasets show 3.56T available, the amount available on old pool.

The old pool was removed from system with a zpool export & drives removed. The new pool was renamed to the old pool's name with export & import.

I tried turning autoexpand on and off and doing "zfs online -e ..." No dice.

How do I get the pool show (allocate?) all available space, which should be close to 7T, not 3.56T?

Output of various commands:

root@vs1:/# zpool get autoexpand
NAME     PROPERTY    VALUE   SOURCE
hs-pool autoexpand on local
root@vs1:/# zpool status
  pool: hs-pool
 state: ONLINE
  scan: none requested
config:

        NAME                                   STATE     READ WRITE CKSUM
        hs-pool                                ONLINE       0     0     0
          mirror-0                             ONLINE       0     0     0
            ata-WDC_WD80EFAX-68KNBN0_VAJAM9ML  ONLINE       0     0     0
            ata-WDC_WD80EFAX-68KNBN0_VAJBD2ML  ONLINE       0     0     0

errors: No known data errors
root@vs1:/# zpool list
NAME      SIZE  ALLOC   FREE  EXPANDSZ   FRAG    CAP  DEDUP  HEALTH  ALTROOT
hs-pool 7.25T 3.46T 3.79T - 29% 47% 1.00x ONLINE -
root@vs1:/# zfs list
NAME USED AVAIL REFER MOUNTPOINT
hs-pool 3.46T 3.56T 96K /hs-pool
hs-pool/backups 85.4G 3.56T 64.7G /hs-pool/backups
hs-pool/downloads 158G 3.56T 158G /hs-pool/downloads
hs-pool/media 2.75T 3.56T 2.75T /hs-pool/media
hs-pool/users 488G 3.56T 488G /hs-pool/users

 

[ph0x]

Size says 7.25 TB with roughly 3.5 TB used and roughly 3.5 TB available.
Are you sure that the former pool was in fact a mirror? Should have been almost full with 2.75 TB media ...

 

[apoc]

Check if you have snapshots that are unexpected. IIRC ZFS send/receive creates snapshots (depending on your mode of operation).
Also have you used compression and/or dedup on the other pool?
This also could explain the differences seen

 

[kamiller42]

Size says 7.25 TB with roughly 3.5 TB used and roughly 3.5 TB available.
Are you sure that the former pool was in fact a mirror? Should have been almost full with 2.75 TB media ...

It was a mirror.

Check if you have snapshots that are unexpected. IIRC ZFS send/receive creates snapshots (depending on your mode of operation).
Also have you used compression and/or dedup on the other pool?
This also could explain the differences seen

I used a snapshot to transfer & then deleted post-transfer.

Guys,

Thank you for responding. I think I'm reading the tool incorrectly, specifically "zfs list". I read AVAIL as total SIZE. When I looked at space using df, I see this.

root@vs1:/# df -h
Filesystem               Size  Used Avail Use% Mounted on
hs-pool                  3.6T     0  3.6T   0% /hs-pool
hs-pool/backups          3.7T   65G  3.6T   2% /hs-pool/backups
hs-pool/downloads        3.8T  158G  3.6T   5% /hs-pool/downloads
hs-pool/media            6.4T  2.8T  3.6T  44% /hs-pool/media
hs-pool/users            4.1T  489G  3.6T  12% /hs-pool/users

I see the new sizes & 3.6TB available across the board. Not understanding why the size varies. (Maybe based on allotment scheme I made, which was 6 years ago.), but I'm happy to see 3.6TB.