X-Git-Url: https://git.proxmox.com/?a=blobdiff_plain;f=PVE%2FDiskmanage.pm;h=55da2e5f285a34b70c5c9e27fa8220866efb8710;hb=d93817829870e2daddcc5d9bd6b86a680b771347;hp=196eee5311e650a047de21904854f861fee0d62e;hpb=8cd6d7e8e3c3a8897f6aba80f91b8da2941eebc2;p=pve-storage.git diff --git a/PVE/Diskmanage.pm b/PVE/Diskmanage.pm index 196eee5..55da2e5 100644 --- a/PVE/Diskmanage.pm +++ b/PVE/Diskmanage.pm @@ -2,10 +2,13 @@ package PVE::Diskmanage; use strict; use warnings; + use PVE::ProcFSTools; use Data::Dumper; use Cwd qw(abs_path); use Fcntl ':mode'; +use File::Basename; +use File::stat; use JSON; use PVE::Tools qw(extract_param run_command file_get_contents file_read_firstline dir_glob_regex dir_glob_foreach trim); @@ -15,9 +18,13 @@ my $ZPOOL = "/sbin/zpool"; my $SGDISK = "/sbin/sgdisk"; my $PVS = "/sbin/pvs"; my $LVS = "/sbin/lvs"; -my $UDEVADM = "/bin/udevadm"; my $LSBLK = "/bin/lsblk"; +sub check_bin { + my ($path) = @_; + return -x $path; +} + sub verify_blockdev_path { my ($rel_path) = @_; @@ -107,7 +114,6 @@ sub get_smart_data { if (defined($type) && $type eq 'ata' && $line =~ m/^([ \d]{2}\d)\s+(\S+)\s+(\S{6})\s+(\d+)\s+(\d+)\s+(\S+)\s+(\S+)\s+(.*)$/) { my $entry = {}; - $entry->{name} = $2 if defined $2; $entry->{flags} = $3 if defined $3; # the +0 makes a number out of the strings @@ -133,8 +139,8 @@ sub get_smart_data { } elsif (defined($type) && $type eq 'text') { $smartdata->{text} = '' if !defined $smartdata->{text}; $smartdata->{text} .= "$line\n"; - # extract wearout from nvme text, allow for decimal values - if ($line =~ m/Percentage Used:.*(\d+(?:\.\d+)?)\%/i) { + # extract wearout from nvme/sas text, allow for decimal values + if ($line =~ m/Percentage Used(?: endurance indicator)?:\s*(\d+(?:\.\d+)?)\%/i) { $smartdata->{wearout} = 100 - $1; } } elsif ($line =~ m/SMART Disabled/) { @@ -156,8 +162,8 @@ sub get_smart_data { return $smartdata; } -sub get_parttype_info() { - my $cmd = [$LSBLK, '--json', '-o', 'path,parttype']; +sub get_lsblk_info() { + my $cmd = [$LSBLK, '--json', '-o', 'path,parttype,fstype']; my $output = ""; my $res = {}; eval { @@ -173,33 +179,35 @@ sub get_parttype_info() { warn "$@\n" if $@; my $list = $parsed->{blockdevices} // []; - foreach my $dev (@$list) { - next if !($dev->{parttype}); - my $type = $dev->{parttype}; - $res->{$type} = [] if !defined($res->{$type}); - push @{$res->{$type}}, $dev->{path}; - } + $res = { map { + $_->{path} => { + parttype => $_->{parttype}, + fstype => $_->{fstype} + } + } @{$list} }; return $res; } my $get_devices_by_partuuid = sub { - my ($parttype_map, $uuids, $res) = @_; + my ($lsblk_info, $uuids, $res) = @_; $res = {} if !defined($res); - foreach my $uuid (sort keys %$uuids) { - map { $res->{$_} = $uuids->{$uuid} } @{$parttype_map->{$uuid}}; + foreach my $dev (sort keys %{$lsblk_info}) { + my $uuid = $lsblk_info->{$dev}->{parttype}; + next if !defined($uuid) || !defined($uuids->{$uuid}); + $res->{$dev} = $uuids->{$uuid}; } return $res; }; sub get_zfs_devices { - my ($parttype_map) = @_; + my ($lsblk_info) = @_; my $res = {}; - return {} if ! -x $ZPOOL; + return {} if !check_bin($ZPOOL); # use zpool and parttype uuid, # because log and cache do not have @@ -224,13 +232,13 @@ sub get_zfs_devices { }; - $res = $get_devices_by_partuuid->($parttype_map, $uuids, $res); + $res = $get_devices_by_partuuid->($lsblk_info, $uuids, $res); return $res; } sub get_lvm_devices { - my ($parttype_map) = @_; + my ($lsblk_info) = @_; my $res = {}; eval { run_command([$PVS, '--noheadings', '--readonly', '-o', 'pv_name'], outfunc => sub{ @@ -250,13 +258,13 @@ sub get_lvm_devices { "e6d6d379-f507-44c2-a23c-238f2a3df928" => 1, }; - $res = $get_devices_by_partuuid->($parttype_map, $uuids, $res); + $res = $get_devices_by_partuuid->($lsblk_info, $uuids, $res); return $res; } sub get_ceph_journals { - my ($parttype_map) = @_; + my ($lsblk_info) = @_; my $res = {}; my $uuids = { @@ -266,7 +274,7 @@ sub get_ceph_journals { 'cafecafe-9b03-4f30-b4c6-b4b80ceff106' => 4, # block }; - $res = $get_devices_by_partuuid->($parttype_map, $uuids, $res); + $res = $get_devices_by_partuuid->($lsblk_info, $uuids, $res); return $res; } @@ -285,13 +293,16 @@ sub get_ceph_volume_infos { my $fields = [ split(';', $line) ]; # lvs syntax is /dev/sdX(Y) where Y is the start (which we do not need) - my ($dev) = $fields->[0] =~ m|^(/dev/[a-z]+)|; + my ($dev) = $fields->[0] =~ m|^(/dev/[a-z]+[^(]*)|; if ($fields->[1] =~ m|^osd-([^-]+)-|) { my $type = $1; # $result autovivification is wanted, to not creating empty hashes if (($type eq 'block' || $type eq 'data') && $fields->[2] =~ m/ceph.osd_id=([^,]+)/) { $result->{$dev}->{osdid} = $1; $result->{$dev}->{bluestore} = ($type eq 'block'); + if ($fields->[2] =~ m/ceph\.encrypted=1/) { + $result->{$dev}->{encrypted} = 1; + } } else { # undef++ becomes '1' (see `perldoc perlop`: Auto-increment) $result->{$dev}->{$type}++; @@ -308,7 +319,7 @@ sub get_udev_info { my $info = ""; my $data = {}; eval { - run_command([$UDEVADM, 'info', '-p', $dev, '--query', 'all'], outfunc => sub { + run_command(['udevadm', 'info', '-p', $dev, '--query', 'all'], outfunc => sub { my ($line) = @_; $info .= "$line\n"; }); @@ -355,9 +366,25 @@ sub get_udev_info { $data->{wwn} = $1; } + if ($info =~ m/^E: DEVLINKS=(.+)$/m) { + my @devlinks = grep(m#^/dev/disk/by-id/(ata|scsi|nvme(?!-eui))#, split (/ /, $1)); + $data->{by_id_link} = $devlinks[0] if defined($devlinks[0]); + } + return $data; } +sub get_sysdir_size { + my ($sysdir) = @_; + + my $size = file_read_firstline("$sysdir/size"); + return if !$size; + + # linux always considers sectors to be 512 bytes, + # independently of real block size + return $size * 512; +} + sub get_sysdir_info { my ($sysdir) = @_; @@ -365,12 +392,7 @@ sub get_sysdir_info { my $data = {}; - my $size = file_read_firstline("$sysdir/size"); - return undef if !$size; - - # linux always considers sectors to be 512 bytes, - # independently of real block size - $data->{size} = $size * 512; + $data->{size} = get_sysdir_size($sysdir) or return; # dir/queue/rotational should be 1 for hdd, 0 for ssd $data->{rotational} = file_read_firstline("$sysdir/queue/rotational") // -1; @@ -387,7 +409,7 @@ sub get_sysdir_info { } sub get_wear_leveling_info { - my ($smartdata, $model) = @_; + my ($smartdata) = @_; my $attributes = $smartdata->{attributes}; if (defined($smartdata->{wearout})) { @@ -396,37 +418,35 @@ sub get_wear_leveling_info { my $wearout; - my $vendormap = { - 'kingston' => 231, - 'samsung' => 177, - 'intel' => 233, - 'sandisk' => 233, - 'crucial' => 202, - 'default' => 233, - }; - - # find target attr id - - my $attrid; - - foreach my $vendor (keys %$vendormap) { - if ($model =~ m/$vendor/i) { - $attrid = $vendormap->{$vendor}; - # found the attribute - last; + # Common register names that represent percentage values of potential + # failure indicators used in drivedb.h of smartmontool's. Order matters, + # as some drives may have multiple definitions + my @wearoutregisters = ( + "Media_Wearout_Indicator", + "SSD_Life_Left", + "Wear_Leveling_Count", + "Perc_Write\/Erase_Ct_BC", + "Perc_Rated_Life_Remain", + "Remaining_Lifetime_Perc", + "Percent_Lifetime_Remain", + "Lifetime_Left", + "PCT_Life_Remaining", + "Lifetime_Remaining", + "Percent_Life_Remaining", + "Percent_Lifetime_Used", + "Perc_Rated_Life_Used" + ); + + # Search for S.M.A.R.T. attributes for known register + foreach my $register (@wearoutregisters) { + last if defined $wearout; + foreach my $attr (@$attributes) { + next if $attr->{name} !~ m/$register/; + $wearout = $attr->{value}; + last; } } - if (!$attrid) { - $attrid = $vendormap->{default}; - } - - foreach my $attr (@$attributes) { - next if $attr->{id} != $attrid; - $wearout = $attr->{value}; - last; - } - return $wearout; } @@ -455,10 +475,12 @@ sub is_iscsi { return 0; } -sub get_disks { - my ($disks, $nosmart) = @_; - my $disklist = {}; +my sub is_ssdlike { + my ($type) = @_; + return $type eq 'ssd' || $type eq 'nvme'; +} +sub mounted_blockdevs { my $mounted = {}; my $mounts = PVE::ProcFSTools::parse_proc_mounts(); @@ -468,19 +490,23 @@ sub get_disks { $mounted->{abs_path($mount->[0])} = $mount->[1]; }; - my $dev_is_mounted = sub { - my ($dev) = @_; - return $mounted->{$dev}; - }; + return $mounted; +} + +sub get_disks { + my ($disks, $nosmart, $include_partitions) = @_; + my $disklist = {}; + + my $mounted = mounted_blockdevs(); - my $parttype_map = get_parttype_info(); + my $lsblk_info = get_lsblk_info(); - my $journalhash = get_ceph_journals($parttype_map); + my $journalhash = get_ceph_journals($lsblk_info); my $ceph_volume_infos = get_ceph_volume_infos(); - my $zfshash = get_zfs_devices($parttype_map); + my $zfshash = get_zfs_devices($lsblk_info); - my $lvmhash = get_lvm_devices($parttype_map); + my $lvmhash = get_lvm_devices($lsblk_info); my $disk_regex = ".*"; if (defined($disks)) { @@ -490,7 +516,7 @@ sub get_disks { die "disks is not a string or array reference\n"; } # we get cciss/c0d0 but need cciss!c0d0 - map { s|cciss/|cciss!| } @$disks; + $_ =~ s|cciss/|cciss!| for @$disks; $disk_regex = "(?:" . join('|', @$disks) . ")"; } @@ -524,6 +550,7 @@ sub get_disks { if ($sysdata->{rotational} == 0) { $type = 'ssd'; + $type = 'nvme' if $dev =~ m/^nvme\d+n\d+$/; $data->{rpm} = 0; } elsif ($sysdata->{rotational} == 1) { if ($data->{rpm} != -1) { @@ -539,25 +566,17 @@ sub get_disks { if (!$nosmart) { eval { - my $smartdata = get_smart_data($devpath, ($type ne 'ssd')); + my $smartdata = get_smart_data($devpath, !is_ssdlike($type)); $health = $smartdata->{health} if $smartdata->{health}; - if ($type eq 'ssd') { + if (is_ssdlike($type)) { # if we have an ssd we try to get the wearout indicator - my $wearval = get_wear_leveling_info($smartdata, $data->{model} || $sysdata->{model}); + my $wearval = get_wear_leveling_info($smartdata); $wearout = $wearval if defined($wearval); } }; } - my $used; - - $used = 'LVM' if $lvmhash->{$devpath}; - - $used = 'mounted' if &$dev_is_mounted($devpath); - - $used = 'ZFS' if $zfshash->{$devpath}; - # we replaced cciss/ with cciss! above # but in the result we need cciss/ again # because the caller might want to check the @@ -580,18 +599,17 @@ sub get_disks { wearout => $wearout, }; + my $by_id_link = $data->{by_id_link}; + $disklist->{$dev}->{by_id_link} = $by_id_link if defined($by_id_link); + my $osdid = -1; my $bluestore = 0; + my $osdencrypted = 0; my $journal_count = 0; my $db_count = 0; my $wal_count = 0; - my $found_partitions; - my $found_lvm; - my $found_mountpoints; - my $found_zfs; - my $found_dm; my $partpath = $devpath; # remove part after last / to @@ -599,65 +617,134 @@ sub get_disks { # e.g. from /dev/cciss/c0d0 get /dev/cciss $partpath =~ s/\/[^\/]+$//; + my $determine_usage = sub { + my ($devpath, $sysdir, $is_partition) = @_; + + return 'LVM' if $lvmhash->{$devpath}; + return 'ZFS' if $zfshash->{$devpath}; + + my $info = $lsblk_info->{$devpath} // {}; + + my $parttype = $info->{parttype}; + if (defined($parttype)) { + return 'BIOS boot' + if $parttype eq '21686148-6449-6e6f-744e-656564454649'; + return 'EFI' + if $parttype eq 'c12a7328-f81f-11d2-ba4b-00a0c93ec93b'; + return 'ZFS reserved' + if $parttype eq '6a945a3b-1dd2-11b2-99a6-080020736631'; + } + + my $fstype = $info->{fstype}; + if (defined($fstype)) { + return "${fstype} (mounted)" if $mounted->{$devpath}; + return "${fstype}"; + } + return 'mounted' if $mounted->{$devpath}; + + return if !$is_partition; + + # for devices, this check is done explicitly later + return 'Device Mapper' if !dir_is_empty("$sysdir/holders"); + + return 'partition'; + }; + + my $collect_ceph_info = sub { + my ($devpath) = @_; + + my $ceph_volume = $ceph_volume_infos->{$devpath} or return; + $journal_count += $ceph_volume->{journal} // 0; + $db_count += $ceph_volume->{db} // 0; + $wal_count += $ceph_volume->{wal} // 0; + if (defined($ceph_volume->{osdid})) { + $osdid = $ceph_volume->{osdid}; + $bluestore = 1 if $ceph_volume->{bluestore}; + $osdencrypted = 1 if $ceph_volume->{encrypted}; + } + + my $result = { %{$ceph_volume} }; + $result->{journals} = delete $result->{journal} + if $result->{journal}; + return $result; + }; + + my $partitions = {}; + dir_glob_foreach("$sysdir", "$dev.+", sub { my ($part) = @_; - $found_partitions = 1; + $partitions->{$part} = $collect_ceph_info->("$partpath/$part"); + my $lvm_based_osd = defined($partitions->{$part}); + + $partitions->{$part}->{devpath} = "$partpath/$part"; + $partitions->{$part}->{parent} = "$devpath"; + $partitions->{$part}->{gpt} = $data->{gpt}; + $partitions->{$part}->{type} = 'partition'; + $partitions->{$part}->{size} = + get_sysdir_size("$sysdir/$part") // 0; + $partitions->{$part}->{used} = + $determine_usage->("$partpath/$part", "$sysdir/$part", 1); + $partitions->{$part}->{osdid} //= -1; + + # Avoid counting twice (e.g. partition on which the LVM for the + # DB OSD resides is present in the $journalhash) + return if $lvm_based_osd; + + # Legacy handling for non-LVM based OSDs - if (my $mp = &$dev_is_mounted("$partpath/$part")) { - $found_mountpoints = 1; + if (my $mp = $mounted->{"$partpath/$part"}) { if ($mp =~ m|^/var/lib/ceph/osd/ceph-(\d+)$|) { $osdid = $1; + $partitions->{$part}->{osdid} = $osdid; } } - if ($lvmhash->{"$partpath/$part"}) { - $found_lvm = 1; - } - - if ($zfshash->{"$partpath/$part"}) { - $found_zfs = 1; - } - if (my $journal_part = $journalhash->{"$partpath/$part"}) { $journal_count++ if $journal_part == 1; $db_count++ if $journal_part == 2; $wal_count++ if $journal_part == 3; $bluestore = 1 if $journal_part == 4; - } - if (!dir_is_empty("$sysdir/$part/holders") && !$found_lvm) { - $found_dm = 1; + $partitions->{$part}->{journals} = 1 if $journal_part == 1; + $partitions->{$part}->{db} = 1 if $journal_part == 2; + $partitions->{$part}->{wal} = 1 if $journal_part == 3; + $partitions->{$part}->{bluestore} = 1 if $journal_part == 4; } }); - if (my $ceph_volume = $ceph_volume_infos->{$devpath}) { - $journal_count += $ceph_volume->{journal} // 0; - $db_count += $ceph_volume->{db} // 0; - $wal_count += $ceph_volume->{wal} // 0; - if (defined($ceph_volume->{osdid})) { - $osdid = $ceph_volume->{osdid}; - $bluestore = 1 if $ceph_volume->{bluestore}; + my $used = $determine_usage->($devpath, $sysdir, 0); + if (!$include_partitions) { + foreach my $part (sort keys %{$partitions}) { + next if $partitions->{$part}->{used} eq 'partition'; + $used //= $partitions->{$part}->{used}; } + } else { + # fstype might be set even if there are partitions, but showing that is confusing + $used = 'partitions' if scalar(keys %{$partitions}); } - - $used = 'mounted' if $found_mountpoints && !$used; - $used = 'LVM' if $found_lvm && !$used; - $used = 'ZFS' if $found_zfs && !$used; - $used = 'Device Mapper' if $found_dm && !$used; - $used = 'partitions' if $found_partitions && !$used; - + $used //= 'partitions' if scalar(keys %{$partitions}); # multipath, software raid, etc. # this check comes in last, to show more specific info # if we have it - $used = 'Device Mapper' if !$used && !dir_is_empty("$sysdir/holders"); + $used //= 'Device Mapper' if !dir_is_empty("$sysdir/holders"); $disklist->{$dev}->{used} = $used if $used; + + $collect_ceph_info->($devpath); + $disklist->{$dev}->{osdid} = $osdid; $disklist->{$dev}->{journals} = $journal_count if $journal_count; $disklist->{$dev}->{bluestore} = $bluestore if $osdid != -1; + $disklist->{$dev}->{osdencrypted} = $osdencrypted if $osdid != -1; $disklist->{$dev}->{db} = $db_count if $db_count; $disklist->{$dev}->{wal} = $wal_count if $wal_count; + + if ($include_partitions) { + foreach my $part (keys %{$partitions}) { + $disklist->{$part} = $partitions->{$part}; + } + } }); return $disklist; @@ -667,18 +754,20 @@ sub get_disks { sub get_partnum { my ($part_path) = @_; - my ($mode, $rdev) = (stat($part_path))[2,6]; + my $st = stat($part_path); - next if !$mode || !S_ISBLK($mode) || !$rdev; - my $major = PVE::Tools::dev_t_major($rdev); - my $minor = PVE::Tools::dev_t_minor($rdev); + die "error detecting block device '$part_path'\n" + if !$st || !$st->mode || !S_ISBLK($st->mode) || !$st->rdev; + + my $major = PVE::Tools::dev_t_major($st->rdev); + my $minor = PVE::Tools::dev_t_minor($st->rdev); my $partnum_path = "/sys/dev/block/$major:$minor/"; my $partnum; $partnum = file_read_firstline("${partnum_path}partition"); - die "Partition does not exists\n" if !defined($partnum); + die "Partition does not exist\n" if !defined($partnum); #untaint and ensure it is a int if ($partnum =~ m/(\d+)/) { @@ -694,9 +783,12 @@ sub get_partnum { sub get_blockdev { my ($part_path) = @_; - my $dev = $1 if $part_path =~ m|^/dev/(.*)$|; - my $link = readlink "/sys/class/block/$dev"; - my $block_dev = $1 if $link =~ m|([^/]*)/$dev$|; + my ($dev, $block_dev); + if ($part_path =~ m|^/dev/(.*)$|) { + $dev = $1; + my $link = readlink "/sys/class/block/$dev"; + $block_dev = $1 if $link =~ m|([^/]*)/$dev$|; + } die "Can't parse parent device\n" if !defined($block_dev); die "No valid block device\n" if index($dev, $block_dev) == -1; @@ -755,4 +847,87 @@ sub append_partition { return $partition; } +my sub strip_dev :prototype($) { + my ($devpath) = @_; + $devpath =~ s|^/dev/||; + return $devpath; +} + +# Check if a disk or any of its partitions has a holder. +# Can also be called with a partition. +# Expected to be called with a result of verify_blockdev_path(). +sub has_holder { + my ($devpath) = @_; + + my $dev = strip_dev($devpath); + + return $devpath if !dir_is_empty("/sys/class/block/${dev}/holders"); + + my $found; + dir_glob_foreach("/sys/block/${dev}", "${dev}.+", sub { + my ($part) = @_; + $found = "/dev/${part}" if !dir_is_empty("/sys/class/block/${part}/holders"); + }); + + return $found; +} + +# Basic check if a disk or any of its partitions is mounted. +# Can also be called with a partition. +# Expected to be called with a result of verify_blockdev_path(). +sub is_mounted { + my ($devpath) = @_; + + my $mounted = mounted_blockdevs(); + + return $devpath if $mounted->{$devpath}; + + my $dev = strip_dev($devpath); + + my $found; + dir_glob_foreach("/sys/block/${dev}", "${dev}.+", sub { + my ($part) = @_; + my $partpath = "/dev/${part}"; + + $found = $partpath if $mounted->{$partpath}; + }); + + return $found; +} + +# Wipes all labels and the first 200 MiB of a disk/partition (or the whole if it is smaller). +# Expected to be called with a result of verify_blockdev_path(). +sub wipe_blockdev { + my ($devpath) = @_; + + my $devname = basename($devpath); + my $dev_size = PVE::Tools::file_get_contents("/sys/class/block/$devname/size"); + + ($dev_size) = $dev_size =~ m|(\d+)|; # untaint $dev_size + die "Couldn't get the size of the device $devname\n" if !defined($dev_size); + + my $size = ($dev_size * 512 / 1024 / 1024); + my $count = ($size < 200) ? $size : 200; + + my $to_wipe = []; + dir_glob_foreach("/sys/class/block/${devname}", "${devname}.+", sub { + my ($part) = @_; + push $to_wipe->@*, "/dev/${part}" if -b "/dev/${part}"; + }); + + if (scalar($to_wipe->$#*) > 0) { + print "found child partitions to wipe: ". join(', ', $to_wipe->@*) ."\n"; + } + push $to_wipe->@*, $devpath; # put actual device last + + print "wiping block device ${devpath}\n"; + + run_command(['wipefs', '--all', $to_wipe->@*], errmsg => "error wiping '${devpath}'"); + + run_command( + ['dd', 'if=/dev/zero', "of=${devpath}", 'bs=1M', 'conv=fdatasync', "count=${count}"], + errmsg => "error wiping '${devpath}'", + ); +} + 1;