The standard output of dysk is a table with a default set of columns and only the “normal looking” filesystems. You can modify it easily.
Columns
You can run dysk --list-cols for the list of all columns, with their aliases (eg ifree for inodes_free).
All columns
| name | default | meaning |
|---|---|---|
| id | mount point id | |
| dev | device id | |
| filesystem | ✓ | filesystem |
| label | label | |
| type | ✓ | filesystem type |
| remote | whether it’s a remote filesystem | |
| disk | ✓ | short tag of the underlying storage identified |
| used | ✓ | cumulated size of the occupied blocks |
| use | ✓ | graphical view of the use share |
| use_percent | percentage of occupied blocks | |
| use_bar | graphical view of the use share, without percentage | |
| free | ✓ | cumulated size of the available blocks |
| free_percent | percentage of available blocks | |
| size | ✓ | size of the volume |
| inodes_free | available inodes | |
| inodes_used | inodes used | |
| inodes | inodes use share, graphical | |
| inodes_use_percent | inodes use share, in percents | |
| inodes_bar | inodes use share, graphical, without percentage | |
| inodes_total | total number of inodes in the filesystem | |
| mount | ✓ | mounting path |
| uuid | filesystem’s UUID | |
| partuuid | partition’s UUID | |
| options | linux mount options | |
| compress | compression algorithm/level |
Choose columns
With the --cols launch argument, shortened as -c, you can change the displayed columns or their order.
The default set is defined for the casual usage of checking the available volumes and their filling level:
With -c all, you may see all available columns (except use and inodes, whose content is already in use_percent+use_bar and inodes_use_percent+inodes_bar), but that’s normally too much for convenience:
The most obvious use of the --cols argument is the explicit definition of the columns to display.
For example dysk -c label+use+size+disk+mount will show the label, use, size, disk, and mount columns, in that order:
All the default columns (see table above) can be inserted with just default.
Here’s adding the label at the start and the device id at the end, with dysk -c label+default+dev:
If the --cols argument starts or ends with + or -, the default set of columns is implied.
To add the device id and the share of inodes used to the default columns, you do dysk -c +dev+inodes:
To prepend the label column before the default ones, use dysk -c label+:
The - sign removes columns.
And adding an already present column moves it to the end (there are never duplicates).
Here’s removing the fs column and moving the type column to the end, with dysk -c -fs+type:
Sort
With the --sort launch argument, shortened as -s, you can specify the order of displayed rows.
The argument’s value must be either a column name, for example dysk -s dev, or a column name and a direction, for example dysk --sort size-desc.
The desc and asc directions can be abbreviated into d and a.
Without a direction, the usual one for the column is used: for example dysk -s size lists the biggest filesystems first.
For example, sorting on the device id:
Or sorting on the remaining free space, in descending order:
Style
Standard output is tailored for a normal terminal.
If you redirect the output, if you want to copy into a document, or if your terminal doesn’t support some characters or ANSI escape sequences, the --color and --ascii launch arguments may come handy:
--color takes yes, no, or auto (the default: colors only when the output is a terminal).
--bar-width sets the width of the bars of the use, use_bar, inodes, and inodes_bar columns (5 by default).
Remote filesystems
dysk reads the usage of remote filesystems (NFS, SMB, etc.) too.
If that’s slow, --remote-stats no skips it (on Linux and Windows): those filesystems are then only listed with -a.
On Linux, reading the usage of a filesystem gives up after 50ms, and the use column then says timeout.
Such a filesystem is only listed with -a.
You may change the delay with --timeout, eg --timeout 2s or --timeout 500ms, or wait as long as needed with --timeout none.
CSV
With the --csv argument, you can ask dysk to output the table in CSV:
dysk --csv > mounts.csv
You may choose the separator with the --csv-separator argument.
Filters, sorting, and column selection work the same as for standard tables so you may do this:
dysk --csv -f 'size>100G' -c remote+default+inodes > mounts.csv
which would give something like this:
