Manual Pages

Table of Contents


na_df - Displays free disk space.


df [ -i | -r | -s | -S | -x ]
[ -h | -k | -m | -g | -t ]
[ -A | -V ]
[ -L ]
[ pathname | aggrname ]


df displays statistics about the amount of free disk space in one or all volumes or aggregates on a node. All sizes are reported in 1024-byte blocks, unless otherwise requested by one of the -h, -k, -m, -g, or -t options.

The pathname parameter is the pathname to a volume. If it is specified, df reports only on the corresponding volume; otherwise, it reports on every online volume. The -V option allows the default scope (volume) to be specified explicitly.

When the -A option is used, then aggrname should instead be the name of an aggregate; when the -A option is used and no aggrname is specified, df reports on every online aggregate. This option displays the space used by the aggregates in the system, including those embedded in traditional volumes.

If the volume being displayed is a FlexCache volume (see na_flexcache(1) ), then the values displayed will be those of the volume being cached. This acts exactly as if the user had issued the df command on the origin node itself. If the remote source volume is unavailable, the relevant values will be displayed as `---'. If a mix of FlexCache and non-FlexCache volumes is being displayed, then the non-FlexCache volumes will display local state.

To view information of the local storage of FlexCache volumes, the -L flag can be used. All flags other than -A are valid in conjunction with -L, as FlexCache operates on a volume level and consequently aggregate information is unavailable. Use of -L does not cause any traffic to the origin node.

For each volume or aggregate, df displays statistics about snapshots on a separate line from statistics about the active file system. The snapshot line reports the amount of space consumed by all the snapshots in the system. Blocks that are referenced by both the active file system and by one or more snapshots are counted only in the active file system line, not in the snapshot line.

If snapshots consume more space than has been reserved for them by the snap reserve command (see na_snap (1)), then the excess space consumed by snapshots is reported as used by the active file system as well as by snapshots. In this case, it may appear that more blocks have been used in total than are actually present in the file system.

The -x option causes df to suppress the display of the snapshot lines. Specifying the option will not affect the display of the active filesystem lines. This will be the case even if the snapshots overflow the snapshot reserve.

With the -r option, df displays the amount of space that has been reserved for overwrites. This is the amount of space set aside in the volume to ensure overwrites to reserved files succeed, multiplied by the value of the fractional_reserve volume option. If there are no snapshots and no block sharing (e.g. deduplication) it is not necessary to set aside space for overwrites and the value displayed will be 0. The reserved space is already counted in the used space, so the -r option can be used to see what portion of the used space represents space reserved for future use. This value will appear in parentheses if the volume is a flexible volume and its storage is not guaranteed; in this case no physical storage has been reserved and the reservation is effectively disabled.

With the -s option, df displays the amount of disk space that has been saved by block sharing (deduplication and file clones) within the volume. With the -S option, df displays the amount of disk space that has been saved by compression and by block sharing (deduplication and file clones) within the volume and also the total space saving because of both block sharing and compression.

The -h option scales the units of each size-related field to be KB, MB, GB, or TB, whichever is most appropriate for the value being displayed. The -k, -m, -g, and -t options scale each size-related field of the output to be expressed in kilobytes, megabytes, gigabytes, or terabytes respectively. Unit values are based on powers of two. For example, one megabyte is equal to 1,048,576 bytes.

With the -i option, df displays statistics on the number of free inodes.


The following example shows file system disk space usage:
  toaster> df
  Filesystem          kbytes   used     avail   capacity Mounted on
  /vol/vol0           4339168  1777824  2561344 41%      /vol/vol0
  /vol/vol0/.snapshot 1084788  956716   128072  88%      /vol/vol0/.snapshot

If snapshots consume more than 100% of the space reserved for them, then either the snapshot reserve should be increased (using snap reserve) or else some of the snapshots should be deleted (using snap delete ). After deleting some snapshots, it may make sense to alter the volume's snapshot schedule (using snap schedule) to reduce the number of snapshots that are kept online.

The following example shows file system inode usage for a specified volume:

  toaster> df -i /vol/vol0
  Filesystem          iused    ifree    %iused  Mounted on
  /vol/vol0           164591   14313    92%     /vol/vol0

You can increase the number of inodes in a file system at any time using the maxfiles command (see maxfiles(1)).

The following example shows disk space usage for aggregate aggr1:

  toaster> df -A aggr1
  Aggregate           kbytes   used     avail   capacity
  aggr1               4339168  1777824  2561344 41%
  aggr1/.snapshot     1084788  956716   128072  88%

The following example shows the statistics of block sharing on volumes.

  toaster> df -s
  Filesystem          used     saved    %saved
  /vol/vol0           2294520  0
  /vol/dense_vol      169708   81996    32%
  /vol/dedup_vol      19640    3620     15%

The disk space savings generated by the shared space is shown in the saved column. The space used plus the space saved would be the total disk space usage, if no space was shared. The %saved is calculated as [saved / (used + saved)].

The following example shows the statistics of compression and block sharing on volumes.

  toaster> df -S
  Filesystem                used       compressed         a-sis     %saved
  /vol/vol0/               21440                0             0         0%
  /vol/vol_1/              52464                0          9008        15%
  /vol/vol_2/              54936            24892             0        31%
  /vol/vol_3/              45204            20984          2184        34%

The disk space savings generated by compression is shown in the compressed column. The disk space savings generated by the shared space is shown in the a-sis column. The space used plus the space compressed plus the space saved by a-sis would be the total disk space usage without compression and block sharing. The %saved is calculated as [(compressed + a-sis) / (used + compressed + a-sis)].


When run from a vfiler context, (e.g. via the vfiler run command), df displays information about only those filesystems that are owned by the concerned vfiler.


na_maxfiles(1), na_rc(5), na_snap(1) na_aggr(1) na_vol(1)


On some NFS clients, the df command does not follow the NFS protocol specification correctly and may display incorrect information about the size of large file systems. Some versions report negative file system sizes; others report a maximum file system size of 2 GB, no matter how large the file system actually is.

Table of Contents