path: root/tools/virt-df

#!/usr/bin/perl -w
# virt-df
# Copyright (C) 2009-2010 Red Hat Inc.
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.

use warnings;
use strict;

use Sys::Guestfs;
use Sys::Guestfs::Lib qw(feature_available);

use Pod::Usage;
use Getopt::Long;
use File::Basename qw(basename);
use POSIX qw(ceil);

use Locale::TextDomain 'libguestfs';

=encoding utf8

=head1 NAME

virt-df - Display free space on virtual filesystems

=head1 SYNOPSIS

 virt-df [--options]

 virt-df [--options] domname

 virt-df [--options] disk.img [disk.img ...]

=head1 DESCRIPTION

C<virt-df> is a command line tool to display free space on virtual
machine filesystems.  Unlike other tools, it doesn't just display the
amount of space allocated to a virtual machine, but can look inside
the virtual machine to see how much space is really being used.

It is like the L<df(1)> command, but for virtual machines, except that
it also works for Windows virtual machines.

If used without any arguments, C<virt-df> checks with libvirt to get a
list of all active and inactive guests, and performs a C<df>-type
operation on each one in turn, printing out the results.

If used with any argument(s), C<virt-df> performs a C<df>-type
operation on either the single named libvirt domain, or on the disk
image(s) listed on the command line (which must all belong to a single
VM).  In this mode (with arguments), C<virt-df> will I<only work for a
single guest>.  If you want to run on multiple guests, then you have
to invoke C<virt-df> multiple times.

Use the C<--csv> option to get a format which can be easily parsed by
other programs.  Other options are mostly similar to standard C<df>
options.  See below for the complete list.

=head1 OPTIONS

=over 4

=cut

my $help;

=item B<--help>

Display brief help.

=cut

my $version;

=item B<--version>

Display version number and exit.

=cut

my $uri;

=item B<--connect URI> | B<-c URI>

If using libvirt, connect to the given I<URI>.  If omitted, then we
connect to the default libvirt hypervisor.

If you specify guest block devices directly, then libvirt is not used
at all.

=cut

my $csv;

=item B<--csv>

Write out the results in CSV format (comma-separated values).  This format
can be imported easily into databases and spreadsheets, but
read L</NOTE ABOUT CSV FORMAT> below.

=cut

my $format;

=item B<--format> raw

Specify the format of disk images given on the command line.  If this
is omitted then the format is autodetected from the content of the
disk image.

If disk images are requested from libvirt, then this program asks
libvirt for this information.  In this case, the value of the format
parameter is ignored.

If working with untrusted raw-format guest disk images, you should
ensure the format is always specified.

=cut

my $human;

=item B<--human-readable> | B<-h>

Print sizes in human-readable format.

You are not allowed to use I<-h> and I<--csv> at the same time.

=cut

my $inodes;

=item B<--inodes> | B<-i>

Print inodes instead of blocks.

=cut

my $one_per_guest;

=item B<--one-per-guest>

Run one libguestfs appliance per guest.  Normally C<virt-df> will
add the disks from several guests to a single libguestfs appliance.

You might use this option in the following circumstances:

=over 4

=item *

If you think an untrusted guest might actively try to exploit the
libguestfs appliance kernel, then this prevents one guest from
interfering with the stats printed for another guest.

=item *

If the kernel has a bug which stops it from accessing a
filesystem in one guest (see for example RHBZ#635373) then
this allows libguestfs to continue and report stats for further
guests.

=back

=cut

my $uuid;

=item B<--uuid>

Print UUIDs instead of names.  This is useful for following
a guest even when the guest is migrated or renamed, or when
two guests happen to have the same name.

Note that only domains that we fetch from libvirt come with UUIDs.
For disk images, we still print the disk image name even when
this option is specified.

=back

=cut

GetOptions ("help|?" => \$help,
            "version" => \$version,
            "connect|c=s" => \$uri,
            "csv" => \$csv,
            "format=s" => \$format,
            "human-readable|human|h" => \$human,
            "inodes|i" => \$inodes,
            "one-per-guest" => \$one_per_guest,
            "uuid" => \$uuid,
    ) or pod2usage (2);
pod2usage (1) if $help;
if ($version) {
    my $g = Sys::Guestfs->new ();
    my %h = $g->version ();
    print "$h{major}.$h{minor}.$h{release}$h{extra}\n";
    exit
}

# RHBZ#600977
die __"virt-df: cannot use -h and --csv options together\n" if $human && $csv;

# RHBZ#635373
#
# Limit the number of devices we will ever add to the appliance.  The
# overall limit in current libguestfs is 25: 26 = number of letters in
# the English alphabet since we are only confident that /dev/sd[a-z]
# will work because of various limits, minus 1 because that may be
# used by the ext2 initial filesystem.
my $max_disks = 25;

# Get the list of domains and block devices.
#
# We can't use Sys::Guestfs::Lib::open_guest here because we want to
# create the libguestfs handle/appliance as few times as possible.
#
# If virt-df is called with no parameters, then run the libvirt
# equivalent of "virsh list --all", get the XML for each domain, and
# get the disk devices.
#
# If virt-df is called with parameters, assume it must either be a
# single disk image filename, a list of disk image filenames, or a
# single libvirt guest name.  Construct disk devices accordingly.

my @domains = ();

if (@ARGV == 0) {               # No params, use libvirt.
    my $conn;

    if ($uri) {
        $conn = Sys::Virt->new (readonly => 1, address => $uri);
    } else {
        $conn = Sys::Virt->new (readonly => 1);
    }

    my @doms = $conn->list_defined_domains ();
    push @doms, $conn->list_domains ();

    # https://bugzilla.redhat.com/show_bug.cgi?id=538041
    @doms = grep { $_->get_id () != 0 } @doms;

    exit 0 unless @doms;

    foreach my $dom (@doms) {
        my @disks = get_disks_from_libvirt ($dom);
        push @domains, { dom => $dom,
                         name => $dom->get_name (),
                         uuid => $dom->get_uuid_string (),
                         disks => \@disks }
    }
} elsif (@ARGV == 1) {          # One param, could be disk image or domname.
    if (-e $ARGV[0]) {
        push @domains, { name => basename ($ARGV[0]),
                         disks => [ [ $ARGV[0], $format ] ] }
    } else {
        my $conn;

        if ($uri) {
            $conn = Sys::Virt->new (readonly => 1, address => $uri);
        } else {
            $conn = Sys::Virt->new (readonly => 1);
        }

        my $dom = $conn->get_domain_by_name ($ARGV[0])
            or die __x("{name} is not the name of a libvirt domain\n",
                       name => $ARGV[0]);
        my @disks = get_disks_from_libvirt ($dom);
        push @domains, { dom => $dom,
                         name => $dom->get_name (),
                         uuid => $dom->get_uuid_string (),
                         disks => \@disks }
    }
} else {                        # >= 2 params, all disk images.
    my @disks = map { [ $_, $format ] } @ARGV;
    push @domains, { name => basename ($ARGV[0]),
                     disks => \@disks }
}

sub get_disks_from_libvirt
{
    my $dom = shift;
    my $xml = $dom->get_xml_description ();

    my $p = XML::XPath->new (xml => $xml);
    my $nodes = $p->find ('//devices/disk');

    my @disks;
    my $node;
    foreach $node ($nodes->get_nodelist) {
        # The filename can be in dev or file attribute, hence:
        my $filename = $p->find ('./source/@dev', $node);
        unless ($filename) {
            $filename = $p->find ('./source/@file', $node);
            next unless $filename;
        }
        $filename = $filename->to_literal;

        # Get the disk format (may not be set).
        my $format = $p->find ('./driver/@type', $node);
        $format = $format->to_literal if $format;

        push @disks, [ $filename, $format ];
    }

    # Code in Sys::Guestfs::Lib dies here if there are no disks at all.

    return @disks;
}

# Sort the domains by name for display.
@domains = sort { $a->{name} cmp $b->{name} } @domains;

# Since we got this far, we're somewhat sure we're going to
# get to print the result, so display the title.
print_title ();

# To minimize the number of times we have to launch the appliance,
# shuffle as many domains together as we can, but not exceeding
# MAX_DISKS per request.  If --one-per-guest was requested then only
# request disks from a single guest each time.
if ($one_per_guest) {
    foreach (@domains) {
        my @request = ( $_ );
        multi_df (@request);
    }
} else {
    while (@domains) {
        my $n = 0; # number of disks added so far
        my @request = ();
        while (@domains) {
            my $c = @{$domains[0]->{disks}};
            if ($c > $max_disks) {
                warn __x("virt-df: ignoring {name}, it has too many disks ({c} > {max})",
                         name => $domains[0]->{name},
                         c => $c, max => $max_disks);
                next;
            }
            last if $n + $c > $max_disks;
            $n += $c;
            push @request, shift (@domains);
        }
        multi_df (@request);
    }
}

sub multi_df
{
    local $_;
    eval {
        my $g = Sys::Guestfs->new ();

        my ($d, $disk);

        foreach $d (@_) {
            foreach $disk (@{$d->{disks}}) {
                my $filename = $disk->[0];
                my $format = $disk->[1];
                my @args = ($filename);
                push @args, readonly => 1;
                push @args, format => $format if defined $format;
                $g->add_drive_opts (@args);
            }
        }

        $g->launch ();
        my $has_lvm2 = feature_available ($g, "lvm2");

        my @devices = $g->list_devices ();
        my @partitions = $g->list_partitions ();

        my $n = 0;
        foreach $d (@_) {
            my $name = $d->{name};
            my $uuid = $d->{uuid};
            my $nr_disks = @{$d->{disks}};

            # Filter LVM to only the devices applying to the original domain.
            my @devs = @devices[$n .. $n+$nr_disks-1];
            $g->lvm_set_filter (\@devs) if $has_lvm2;

            # Find which whole devices (RHBZ#590167), partitions and LVs
            # contain mountable filesystems.  Stat those which are
            # mountable, and ignore the others.
            foreach (@devs) {
                try_df ($name, $uuid, $g, $_, canonical_dev ($_, $n));
            }
            foreach (filter_partitions (\@devs, @partitions)) {
                try_df ($name, $uuid, $g, $_, canonical_dev ($_, $n));
            }
            if ($has_lvm2) {
                foreach ($g->lvs ()) {
                    try_df ($name, $uuid, $g, $_);
                }
            }

            $n += $nr_disks;
        }
    };
    warn if $@;
}

sub filter_partitions
{
    my $devs = shift;
    my @devs = @$devs;
    my @r;

    foreach my $p (@_) {
        foreach my $d (@devs) {
            if ($p =~ /^$d\d/) {
                push @r, $p;
                last;
            }
        }
    }

    return @r;
}

# Calculate the canonical name for a device.
# eg: /dev/vdb1 when offset = 1
#     => canonical name is /dev/sda1
sub canonical_dev
{
    local $_;
    my $dev = shift;
    my $offset = shift;

    return $dev unless $dev =~ m{^/dev/.d([a-z])(\d*)$};
    my $disk = $1;
    my $partnum = $2;

    $disk = chr (ord ($disk) - $offset);

    return "/dev/sd$disk$partnum"
}

sub try_df
{
    local $_;
    my $domname = shift;
    my $domuuid = shift;
    my $g = shift;
    my $dev = shift;
    my $display = shift || $dev;

    my %stat;
    eval {
        $g->mount_ro ($dev, "/");
        %stat = $g->statvfs ("/");
    };
    if (!$@) {
        print_stat ($domname, $domuuid, $display, \%stat);
    }
    $g->umount_all ();
}

sub print_stat
{
    my $domname = shift;
    my $domuuid = shift;
    my $dev = shift;
    my $stat = shift;

    my @cols;
    if (!$uuid || !defined $domuuid) {
        push @cols, $domname;
    } else {
        push @cols, $domuuid;
    }
    push @cols, $dev;

    if (!$inodes) {
        my $bsize = $stat->{bsize};	# block size
        my $blocks = $stat->{blocks};	# total number of blocks
        my $bfree = $stat->{bfree};	# blocks free (total)
        my $bavail = $stat->{bavail};	# blocks free (for non-root users)

        my $factor = $bsize / 1024;

        push @cols, $blocks*$factor;	# total 1K blocks
        push @cols, ($blocks-$bfree)*$factor; # total 1K blocks used
        push @cols, $bavail*$factor;	# total 1K blocks available

        push @cols, 100.0 - 100.0 * $bfree / $blocks;

        if ($human) {
            $cols[2] = human_size ($cols[2]);
            $cols[3] = human_size ($cols[3]);
            $cols[4] = human_size ($cols[4]);
        }
    } else {
        my $files = $stat->{files};	# total number of inodes
        my $ffree = $stat->{ffree};	# inodes free (total)
        my $favail = $stat->{favail};	# inodes free (for non-root users)

        push @cols, $files;
        push @cols, $files-$ffree;
        push @cols, $ffree;

        push @cols, 100.0 - 100.0 * $ffree / $files;
    }

    print_cols (@cols);
}

sub print_title
{
    my @cols = (__"Virtual Machine", __"Filesystem");
    if (!$inodes) {
        if (!$human) {
            push @cols, __"1K-blocks";
        } else {
            push @cols, __"Size";
        }
        push @cols, __"Used";
        push @cols, __"Available";
        push @cols, __"Use%";
    } else {
        push @cols, __"Inodes";
        push @cols, __"IUsed";
        push @cols, __"IFree";
        push @cols, __"IUse%";
    }

    if (!$csv) {
        # ignore $cols[0] in this mode
        printf "%-36s%10s %10s %10s %5s\n",
          $cols[1], $cols[2], $cols[3], $cols[4], $cols[5];
    } else {
        # Columns don't need special CSV quoting.
        print (join (",", @cols), "\n");
    }
}

sub print_cols
{
    if (!$csv) {
        my $label = sprintf "%s:%s", $_[0], $_[1];

        printf ("%-36s", $label);
        print "\n"," "x36 if length ($label) > 36;

        # Use 'ceil' on the percentage in order to emulate
        # what df itself does.
        my $percent = sprintf "%3d%%", ceil($_[5]);

        printf ("%10s %10s %10s %5s\n", $_[2], $_[3], $_[4], $percent);
    } else {
        # Need to quote libvirt domain and filesystem.
        my $dom = shift;
        my $fs = shift;
        print csv_quote($dom), ",", csv_quote($fs), ",";
        printf ("%d,%d,%d,%.1f%%\n", @_);
    }
}

# Convert a number of 1K blocks to a human-readable number.
sub human_size
{
    local $_ = shift;

    if ($_ < 1024) {
        sprintf "%dK", $_;
    } elsif ($_ < 1024 * 1024) {
        sprintf "%.1fM", ($_ / 1024);
    } else {
        sprintf "%.1fG", ($_ / 1024 / 1024);
    }
}

# Quote field for CSV without using an external module.
sub csv_quote
{
    local $_ = shift;

    my $needs_quoting = /[ ",\n\0]/;
    return $_ unless $needs_quoting;

    my $i;
    my $out = '"';
    for ($i = 0; $i < length; ++$i) {
        my $c = substr $_, $i, 1;
        if ($c eq '"') {
            $out .= '""';
        } elsif ($c eq '\0') {
            $out .= '"0';
        } else {
            $out .= $c;
        }
    }
    $out .= '"';

    return $out;
}

=head1 NOTE ABOUT CSV FORMAT

Comma-separated values (CSV) is a deceptive format.  It I<seems> like
it should be easy to parse, but it is definitely not easy to parse.

Myth: Just split fields at commas.  Reality: This does I<not> work
reliably.  This example has two columns:

 "foo,bar",baz

Myth: Read the file one line at a time.  Reality: This does I<not>
work reliably.  This example has one row:

 "foo
 bar",baz

For shell scripts, use C<csvtool> (L<http://merjis.com/developers/csv>
also packaged in major Linux distributions).

For other languages, use a CSV processing library (eg. C<Text::CSV>
for Perl or Python's built-in csv library).

Most spreadsheets and databases can import CSV directly.

=head1 SHELL QUOTING

Libvirt guest names can contain arbitrary characters, some of which
have meaning to the shell such as C<#> and space.  You may need to
quote or escape these characters on the command line.  See the shell
manual page L<sh(1)> for details.

=head1 SEE ALSO

L<guestfs(3)>,
L<guestfish(1)>,
L<Sys::Guestfs(3)>,
L<Sys::Guestfs::Lib(3)>,
L<Sys::Virt(3)>,
L<http://libguestfs.org/>.

=head1 AUTHOR

Richard W.M. Jones L<http://people.redhat.com/~rjones/>

=head1 COPYRIGHT

Copyright (C) 2009-2010 Red Hat Inc.

This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.