Aggregate-Volume used space mismatch

[ct_ts]

Hi,

We have a problem where used space on our aggregate does not equal the total sum of used space for all volumes.

Filer1> df -Ag
Aggregate                      total              used      avail capacity
aggr0                               9699GB     6042GB     3657GB      62%
aggr0/.snapshot            0GB        0GB        0GB     ---%

Filter1> df -sg
Filesystem                      used      saved       %saved
/vol/vol0/                           5GB        0GB           0%
/vol/vol_vmware_nfs/     8456GB     2058GB          20%

As you can see there is approx 8.4TB of space used by all (thin provisioned) volumes, but only 6TB used on the aggregate.

Even taking away the 2TB of de-dupe savings, there is still nearly 500GB that appears to have gone "missing". Can anyone explain what might be going on here?

I have checked and there are no snapshots currently on the volume.

Thanks,

[Naveenpusuluru]

If you want to go for exact caliculation, go for mb or kb level. Then only you will get exact caliculation. NetApp will prefer round figures. If you have a 1GB volume and that is used 900MB but if you check t with df -g it will show u as 0 only.

df -Am

df -m

[aborzenkov]

Also post output of "aggr show_space".

[ct_ts]

Hi, thanks for the replies. Here is the output. It looks like aggr show_space shows the actual used space of the volume vol_vmware_nfs. Can you help me understand why there is such a difference between the two outputs (9.2TB with df-m and approx 6.6TB with aggr_show_space?)

Aggregate                total       used      avail capacity
aggr0                9932067MB  6757124MB  3174943MB      68%
aggr0/.snapshot            0MB        0MB        0MB     ---%
Filer1> df -m
Filesystem               total       used      avail capacity  Mounted on
/vol/vol0/            157593MB     6106MB   151487MB       4%  /vol/vol0/
/vol/vol0/.snapshot     8294MB      293MB     8001MB       4%  /vol/vol0/.snapshot
/vol/vol_vmware_nfs/ 10485760MB  9200257MB  1285502MB      88%  /vol/vol_vmware_nfs/
/vol/vol_vmware_nfs/.snapshot        0MB   507180MB        0MB     ---%  /vol/vol_vmware_nfs/.snapshot

Filer1> aggr show_space
Aggregate 'aggr0'

    Total space    WAFL reserve    Snap reserve    Usable space       BSR NVLOG           A-SIS          Smtape
  11300486144KB    1130048612KB             0KB   10170437532KB             0KB     103664352KB             0KB

Space allocated to volumes in the aggregate

Volume                          Allocated            Used       Guarantee
vol0                          171898376KB       6000312KB          volume
vol_vmware_nfs               6643458956KB    6607153892KB            none

Aggregate                       Allocated            Used           Avail
Total space                  6815357332KB    6613154204KB    3250943096KB
Snap reserve                          0KB             0KB             0KB
WAFL reserve                 1130048612KB     123729496KB    1006319116KB

[aborzenkov]

My best guess is that this is deduplication effect. I.e. logically your files require about 9000GB on volume; but actual space consumption on aggregate level is reduced due to deduplication.