use PVE::CGroup;
use PVE::DataCenterConfig;
use PVE::Exception qw(raise raise_param_exc);
+use PVE::Format qw(render_duration render_bytes);
use PVE::GuestHelpers qw(safe_string_ne safe_num_ne safe_boolean_ne);
use PVE::INotify;
use PVE::JSONSchema qw(get_standard_option parse_property_string);
use PVE::QemuServer::Cloudinit;
use PVE::QemuServer::CGroup;
use PVE::QemuServer::CPUConfig qw(print_cpu_device get_cpu_options);
-use PVE::QemuServer::Drive qw(is_valid_drivename drive_is_cloudinit drive_is_cdrom parse_drive print_drive);
+use PVE::QemuServer::Drive qw(is_valid_drivename drive_is_cloudinit drive_is_cdrom drive_is_read_only parse_drive print_drive);
use PVE::QemuServer::Machine;
use PVE::QemuServer::Memory;
use PVE::QemuServer::Monitor qw(mon_cmd);
cpuunits => {
optional => 1,
type => 'integer',
- description => "CPU weight for a VM.",
- verbose_description => "CPU weight for a VM. Argument is used in the kernel fair scheduler. The larger the number is, the more CPU time this VM gets. Number is relative to weights of all the other running VMs.",
+ description => "CPU weight for a VM, will be clamped to [1, 10000] in cgroup v2.",
+ verbose_description => "CPU weight for a VM. Argument is used in the kernel fair scheduler."
+ ." The larger the number is, the more CPU time this VM gets. Number is relative to"
+ ." weights of all the other running VMs.",
minimum => 2,
maximum => 262144,
- default => 1024,
+ default => 'cgroup v1: 1024, cgroup v2: 100',
},
memory => {
optional => 1,
description => {
optional => 1,
type => 'string',
- description => "Description for the VM. Only used on the configuration web interface. This is saved as comment inside the configuration file.",
+ description => "Description for the VM. Shown in the web-interface VM's summary."
+ ." This is saved as comment inside the configuration file.",
+ maxLength => 1024 * 8,
},
ostype => {
optional => 1,
$confdesc->{"numa$i"} = $numadesc;
}
-my $nic_model_list = ['rtl8139', 'ne2k_pci', 'e1000', 'pcnet', 'virtio',
- 'ne2k_isa', 'i82551', 'i82557b', 'i82559er', 'vmxnet3',
- 'e1000-82540em', 'e1000-82544gc', 'e1000-82545em'];
+my $nic_model_list = [
+ 'e1000',
+ 'e1000-82540em',
+ 'e1000-82544gc',
+ 'e1000-82545em',
+ 'e1000e',
+ 'i82551',
+ 'i82557b',
+ 'i82559er',
+ 'ne2k_isa',
+ 'ne2k_pci',
+ 'pcnet',
+ 'rtl8139',
+ 'virtio',
+ 'vmxnet3',
+];
my $nic_model_list_txt = join(' ', sort @$nic_model_list);
my $net_fmt_bridge_descr = <<__EOD__;
The special string 'dhcp' can be used for IP addresses to use DHCP, in which case no explicit
gateway should be provided.
-For IPv6 the special string 'auto' can be used to use stateless autoconfiguration.
+For IPv6 the special string 'auto' can be used to use stateless autoconfiguration. This requires
+cloud-init 19.4 or newer.
If cloud-init is enabled and neither an IPv4 nor an IPv6 address is specified, it defaults to using
dhcp on IPv4.
return "usb-kbd,id=keyboard,bus=ehci.0,port=2";
}
+my sub get_drive_id {
+ my ($drive) = @_;
+ return "$drive->{interface}$drive->{index}";
+}
+
sub print_drivedevice_full {
my ($storecfg, $conf, $vmid, $drive, $bridges, $arch, $machine_type) = @_;
my $device = '';
my $maxdev = 0;
- my $drive_id = "$drive->{interface}$drive->{index}";
+ my $drive_id = get_drive_id($drive);
if ($drive->{interface} eq 'virtio') {
my $pciaddr = print_pci_addr("$drive_id", $bridges, $arch, $machine_type);
$device = "virtio-blk-pci,drive=drive-$drive_id,id=${drive_id}${pciaddr}";
}
}
- if (!$conf->{scsihw} || ($conf->{scsihw} =~ m/^lsi/)){
+ if (!$conf->{scsihw} || $conf->{scsihw} =~ m/^lsi/ || $conf->{scsihw} eq 'pvscsi') {
$device = "scsi-$devicetype,bus=$controller_prefix$controller.0,scsi-id=$unit";
} else {
$device = "scsi-$devicetype,bus=$controller_prefix$controller.0,channel=0,scsi-id=0"
}
sub print_drive_commandline_full {
- my ($storecfg, $vmid, $drive) = @_;
+ my ($storecfg, $vmid, $drive, $pbs_name, $io_uring) = @_;
my $path;
my $volid = $drive->{file};
- my $format;
+ my $format = $drive->{format};
+ my $drive_id = get_drive_id($drive);
+
+ my ($storeid, $volname) = PVE::Storage::parse_volume_id($volid, 1);
+ my $scfg = $storeid ? PVE::Storage::storage_config($storecfg, $storeid) : undef;
if (drive_is_cdrom($drive)) {
$path = get_iso_path($storecfg, $vmid, $volid);
+ die "$drive_id: cannot back cdrom drive with PBS snapshot\n" if $pbs_name;
} else {
- my ($storeid, $volname) = PVE::Storage::parse_volume_id($volid, 1);
if ($storeid) {
$path = PVE::Storage::path($storecfg, $volid);
- my $scfg = PVE::Storage::storage_config($storecfg, $storeid);
- $format = qemu_img_format($scfg, $volname);
+ $format //= qemu_img_format($scfg, $volname);
} else {
$path = $volid;
- $format = "raw";
+ $format //= "raw";
}
}
+ my $is_rbd = $path =~ m/^rbd:/;
+
my $opts = '';
- my @qemu_drive_options = qw(heads secs cyls trans media format cache rerror werror aio discard);
+ my @qemu_drive_options = qw(heads secs cyls trans media cache rerror werror aio discard);
foreach my $o (@qemu_drive_options) {
$opts .= ",$o=$drive->{$o}" if defined($drive->{$o});
}
}
}
- $opts .= ",format=$format" if $format && !$drive->{format};
+ if ($pbs_name) {
+ $format = "rbd" if $is_rbd;
+ die "$drive_id: Proxmox Backup Server backed drive cannot auto-detect the format\n"
+ if !$format;
+ $opts .= ",format=alloc-track,file.driver=$format";
+ } elsif ($format) {
+ $opts .= ",format=$format";
+ }
my $cache_direct = 0;
if (my $cache = $drive->{cache}) {
$cache_direct = $cache =~ /^(?:off|none|directsync)$/;
- } elsif (!drive_is_cdrom($drive)) {
+ } elsif (!drive_is_cdrom($drive) && !($scfg && $scfg->{type} eq 'btrfs' && !$scfg->{nocow})) {
$opts .= ",cache=none";
$cache_direct = 1;
}
- # aio native works only with O_DIRECT
+ # io_uring with cache mode writeback or writethrough on krbd will hang...
+ my $rbd_no_io_uring = $scfg && $scfg->{type} eq 'rbd' && $scfg->{krbd} && !$cache_direct;
+
+ # io_uring with cache mode writeback or writethrough on LVM will hang, without cache only
+ # sometimes, just plain disable...
+ my $lvm_no_io_uring = $scfg && $scfg->{type} eq 'lvm';
+
if (!$drive->{aio}) {
- if($cache_direct) {
- $opts .= ",aio=native";
+ if ($io_uring && !$rbd_no_io_uring && !$lvm_no_io_uring) {
+ # io_uring supports all cache modes
+ $opts .= ",aio=io_uring";
} else {
- $opts .= ",aio=threads";
+ # aio native works only with O_DIRECT
+ if($cache_direct) {
+ $opts .= ",aio=native";
+ } else {
+ $opts .= ",aio=threads";
+ }
}
}
# This used to be our default with discard not being specified:
$detectzeroes = 'on';
}
- $opts .= ",detect-zeroes=$detectzeroes" if $detectzeroes;
+
+ # note: 'detect-zeroes' works per blockdev and we want it to persist
+ # after the alloc-track is removed, so put it on 'file' directly
+ my $dz_param = $pbs_name ? "file.detect-zeroes" : "detect-zeroes";
+ $opts .= ",$dz_param=$detectzeroes" if $detectzeroes;
+ }
+
+ if ($pbs_name) {
+ $opts .= ",backing=$pbs_name";
+ $opts .= ",auto-remove=on";
}
- my $pathinfo = $path ? "file=$path," : '';
+ # my $file_param = $pbs_name ? "file.file.filename" : "file";
+ my $file_param = "file";
+ if ($pbs_name) {
+ # non-rbd drivers require the underlying file to be a seperate block
+ # node, so add a second .file indirection
+ $file_param .= ".file" if !$is_rbd;
+ $file_param .= ".filename";
+ }
+ my $pathinfo = $path ? "$file_param=$path," : '';
return "${pathinfo}if=none,id=drive-$drive->{interface}$drive->{index}$opts";
}
+sub print_pbs_blockdev {
+ my ($pbs_conf, $pbs_name) = @_;
+ my $blockdev = "driver=pbs,node-name=$pbs_name,read-only=on";
+ $blockdev .= ",repository=$pbs_conf->{repository}";
+ $blockdev .= ",snapshot=$pbs_conf->{snapshot}";
+ $blockdev .= ",archive=$pbs_conf->{archive}";
+ $blockdev .= ",keyfile=$pbs_conf->{keyfile}" if $pbs_conf->{keyfile};
+ return $blockdev;
+}
+
sub print_netdevice_full {
my ($vmid, $conf, $net, $netid, $bridges, $use_old_bios_files, $arch, $machine_type) = @_;
$romfile = 'pxe-virtio.rom';
} elsif ($device eq 'e1000') {
$romfile = 'pxe-e1000.rom';
+ } elsif ($device eq 'e1000e') {
+ $romfile = 'pxe-e1000e.rom';
} elsif ($device eq 'ne2k') {
$romfile = 'pxe-ne2k_pci.rom';
} elsif ($device eq 'pcnet') {
if ($conf->{template}) {
# check if any base image is still used by a linked clone
- PVE::QemuConfig->foreach_volume($conf, sub {
+ PVE::QemuConfig->foreach_volume_full($conf, { include_unused => 1 }, sub {
my ($ds, $drive) = @_;
return if drive_is_cdrom($drive);
});
}
- # only remove disks owned by this VM (referenced in the config)
- PVE::QemuConfig->foreach_volume_full($conf, { include_unused => 1 }, sub {
+ my $remove_owned_drive = sub {
my ($ds, $drive) = @_;
return if drive_is_cdrom($drive, 1);
eval { PVE::Storage::vdisk_free($storecfg, $volid) };
warn "Could not remove disk '$volid', check manually: $@" if $@;
- });
+ };
+
+ # only remove disks owned by this VM (referenced in the config)
+ my $include_opts = {
+ include_unused => 1,
+ extra_keys => ['vmstate'],
+ };
+ PVE::QemuConfig->foreach_volume_full($conf, $include_opts, $remove_owned_drive);
+
+ for my $snap (values %{$conf->{snapshots}}) {
+ next if !defined($snap->{vmstate});
+ my $drive = PVE::QemuConfig->parse_volume('vmstate', $snap->{vmstate}, 1);
+ next if !defined($drive);
+ $remove_owned_drive->('vmstate', $drive);
+ }
if ($purge_unreferenced) { # also remove unreferenced disk
- my $vmdisks = PVE::Storage::vdisk_list($storecfg, undef, $vmid);
+ my $vmdisks = PVE::Storage::vdisk_list($storecfg, undef, $vmid, undef, 'images');
PVE::Storage::foreach_volid($vmdisks, sub {
my ($volid, $sid, $volname, $d) = @_;
eval { PVE::Storage::vdisk_free($storecfg, $volid) };
return if !$sid;
# check if storage is available on both nodes
- my $scfg = PVE::Storage::storage_check_node($storecfg, $sid);
- PVE::Storage::storage_check_node($storecfg, $sid, $node);
+ my $scfg = PVE::Storage::storage_check_enabled($storecfg, $sid);
+ PVE::Storage::storage_check_enabled($storecfg, $sid, $node);
+
+ my ($vtype) = PVE::Storage::parse_volname($storecfg, $volid);
+
+ die "$volid: content type '$vtype' is not available on storage '$sid'\n"
+ if !$scfg->{content}->{$vtype};
});
}
my $conf = PVE::QemuConfig->load_config($vmid);
- my $d = { vmid => $vmid };
- $d->{pid} = $list->{$vmid}->{pid};
+ my $d = { vmid => int($vmid) };
+ $d->{pid} = int($list->{$vmid}->{pid}) if $list->{$vmid}->{pid};
# fixme: better status?
$d->{status} = $list->{$vmid}->{pid} ? 'running' : 'stopped';
$d->{diskread} = 0;
$d->{diskwrite} = 0;
- $d->{template} = PVE::QemuConfig->is_template($conf);
+ $d->{template} = 1 if PVE::QemuConfig->is_template($conf);
$d->{serial} = 1 if conf_has_serial($conf);
$d->{lock} = $conf->{lock} if $conf->{lock};
$d->{netin} += $netdev->{$dev}->{transmit};
if ($full) {
- $d->{nics}->{$dev}->{netout} = $netdev->{$dev}->{receive};
- $d->{nics}->{$dev}->{netin} = $netdev->{$dev}->{transmit};
+ $d->{nics}->{$dev}->{netout} = int($netdev->{$dev}->{receive});
+ $d->{nics}->{$dev}->{netin} = int($netdev->{$dev}->{transmit});
}
}
aarch64 => 'virt',
};
+sub get_installed_machine_version {
+ my ($kvmversion) = @_;
+ $kvmversion = kvm_user_version() if !defined($kvmversion);
+ $kvmversion =~ m/^(\d+\.\d+)/;
+ return $1;
+}
+
+sub windows_get_pinned_machine_version {
+ my ($machine, $base_version, $kvmversion) = @_;
+
+ my $pin_version = $base_version;
+ if (!defined($base_version) ||
+ !PVE::QemuServer::Machine::can_run_pve_machine_version($base_version, $kvmversion)
+ ) {
+ $pin_version = get_installed_machine_version($kvmversion);
+ }
+ if (!$machine || $machine eq 'pc') {
+ $machine = "pc-i440fx-$pin_version";
+ } elsif ($machine eq 'q35') {
+ $machine = "pc-q35-$pin_version";
+ } elsif ($machine eq 'virt') {
+ $machine = "virt-$pin_version";
+ } else {
+ warn "unknown machine type '$machine', not touching that!\n";
+ }
+
+ return $machine;
+}
+
sub get_vm_machine {
my ($conf, $forcemachine, $arch, $add_pve_version, $kvmversion) = @_;
# layout which confuses windows quite a bit and may result in various regressions..
# see: https://lists.gnu.org/archive/html/qemu-devel/2021-02/msg08484.html
if (windows_version($conf->{ostype})) {
- my $pin_version = '5.1';
- if (!PVE::QemuServer::Machine::can_run_pve_machine_version($pin_version, $kvmversion)) {
- $kvmversion =~ m/^(\d+\.\d+)/;
- $pin_version = $1;
- }
- if (!$machine || $machine eq 'pc') {
- $machine = "pc-i440fx-$pin_version";
- } elsif ($machine eq 'q35') {
- $machine = "pc-q35-$pin_version";
- }
+ $machine = windows_get_pinned_machine_version($machine, '5.1', $kvmversion);
}
$arch //= 'x86_64';
$machine ||= $default_machines->{$arch};
}
}
- if ($add_pve_version && $machine !~ m/\+pve\d+$/) {
+ if ($add_pve_version && $machine !~ m/\+pve\d+?(?:\.pxe)?$/) {
+ my $is_pxe = $machine =~ m/^(.*?)\.pxe$/;
+ $machine = $1 if $is_pxe;
+
# for version-pinned machines that do not include a pve-version (e.g.
# pc-q35-4.1), we assume 0 to keep them stable in case we bump
$machine .= '+pve0';
+
+ $machine .= '.pxe' if $is_pxe;
}
return $machine;
$qemu_cmd,
'-machine', $default_machine,
'-display', 'none',
- '-chardev', "socket,id=qmp,path=/var/run/qemu-server/$fakevmid.qmp,server,nowait",
+ '-chardev', "socket,id=qmp,path=/var/run/qemu-server/$fakevmid.qmp,server=on,wait=off",
'-mon', 'chardev=qmp,mode=control',
'-pidfile', $pidfile,
'-S', '-daemonize'
return \@flags;
}
+my sub get_cpuunits {
+ my ($conf) = @_;
+ return $conf->{cpuunits} // (PVE::CGroup::cgroup_mode() == 2 ? 100 : 1024);
+}
sub config_to_command {
- my ($storecfg, $vmid, $conf, $defaults, $forcemachine, $forcecpu) = @_;
+ my ($storecfg, $vmid, $conf, $defaults, $forcemachine, $forcecpu,
+ $pbs_backing) = @_;
my $cmd = [];
my $globalFlags = [];
my $use_old_bios_files = undef;
($use_old_bios_files, $machine_type) = qemu_use_old_bios_files($machine_type);
- my $cpuunits = defined($conf->{cpuunits}) ?
- $conf->{cpuunits} : $defaults->{cpuunits};
+ my $cpuunits = get_cpuunits($conf);
push @$cmd, $kvm_binary;
my $use_virtio = 0;
my $qmpsocket = PVE::QemuServer::Helpers::qmp_socket($vmid);
- push @$cmd, '-chardev', "socket,id=qmp,path=$qmpsocket,server,nowait";
+ push @$cmd, '-chardev', "socket,id=qmp,path=$qmpsocket,server=on,wait=off";
push @$cmd, '-mon', "chardev=qmp,mode=control";
if (min_version($machine_version, 2, 12)) {
die "uefi base image '$ovmf_code' not found\n" if ! -f $ovmf_code;
my ($path, $format);
+ my $read_only_str = '';
if (my $efidisk = $conf->{efidisk0}) {
my $d = parse_drive('efidisk0', $efidisk);
my ($storeid, $volname) = PVE::Storage::parse_volume_id($d->{file}, 1);
die "efidisk format must be specified\n"
if !defined($format);
}
+
+ $read_only_str = ',readonly=on' if drive_is_read_only($conf, $d);
} else {
warn "no efidisk configured! Using temporary efivars disk.\n";
$path = "/tmp/$vmid-ovmf.fd";
$size_str = ",size=" . (-s $ovmf_vars);
}
- push @$cmd, '-drive', "if=pflash,unit=0,format=raw,readonly,file=$ovmf_code";
- push @$cmd, '-drive', "if=pflash,unit=1,format=$format,id=drive-efidisk0$size_str,file=$path";
+ # SPI flash does lots of read-modify-write OPs, without writeback this gets really slow #3329
+ my $cache = "";
+ if ($path =~ m/^rbd:/) {
+ $cache = ',cache=writeback';
+ $path .= ':rbd_cache_policy=writeback'; # avoid write-around, we *need* to cache writes too
+ }
+
+ push @$cmd, '-drive', "if=pflash,unit=0,format=raw,readonly=on,file=$ovmf_code";
+ push @$cmd, '-drive', "if=pflash,unit=1$cache,format=$format,id=drive-efidisk0$size_str,file=${path}${read_only_str}";
}
# load q35 config
if (my $path = $conf->{"serial$i"}) {
if ($path eq 'socket') {
my $socket = "/var/run/qemu-server/${vmid}.serial$i";
- push @$devices, '-chardev', "socket,id=serial$i,path=$socket,server,nowait";
+ push @$devices, '-chardev', "socket,id=serial$i,path=$socket,server=on,wait=off";
# On aarch64, serial0 is the UART device. Qemu only allows
# connecting UART devices via the '-serial' command line, as
# the device has a fixed slot on the hardware...
push @$devices, '-device', print_vga_device(
$conf, $vga, $arch, $machine_version, $machine_type, undef, $qxlnum, $bridges);
my $socket = PVE::QemuServer::Helpers::vnc_socket($vmid);
- push @$cmd, '-vnc', "unix:$socket,password";
+ push @$cmd, '-vnc', "unix:$socket,password=on";
} else {
push @$cmd, '-vga', 'none' if $vga->{type} eq 'none';
push @$cmd, '-nographic';
if ($guest_agent->{enabled}) {
my $qgasocket = PVE::QemuServer::Helpers::qmp_socket($vmid, 1);
- push @$devices, '-chardev', "socket,path=$qgasocket,server,nowait,id=qga0";
+ push @$devices, '-chardev', "socket,path=$qgasocket,server=on,wait=off,id=qga0";
if (!$guest_agent->{type} || $guest_agent->{type} eq 'virtio') {
my $pciaddr = print_pci_addr("qga0", $bridges, $arch, $machine_type);
my ($ds, $drive) = @_;
if (PVE::Storage::parse_volume_id($drive->{file}, 1)) {
+ check_volume_storage_type($storecfg, $drive->{file});
push @$vollist, $drive->{file};
}
$ahcicontroller->{$controller}=1;
}
- my $drive_cmd = print_drive_commandline_full($storecfg, $vmid, $drive);
- $drive_cmd .= ',readonly' if PVE::QemuConfig->is_template($conf);
+ my $pbs_conf = $pbs_backing->{$ds};
+ my $pbs_name = undef;
+ if ($pbs_conf) {
+ $pbs_name = "drive-$ds-pbs";
+ push @$devices, '-blockdev', print_pbs_blockdev($pbs_conf, $pbs_name);
+ }
+
+ my $drive_cmd = print_drive_commandline_full(
+ $storecfg, $vmid, $drive, $pbs_name, min_version($kvmver, 6, 0));
+
+ # extra protection for templates, but SATA and IDE don't support it..
+ $drive_cmd .= ',readonly=on' if drive_is_read_only($conf, $drive);
push @$devices, '-drive',$drive_cmd;
push @$devices, '-device', print_drivedevice_full(
print "activating and using '$vmstate' as vmstate\n";
}
+ if (PVE::QemuConfig->is_template($conf)) {
+ # needed to workaround base volumes being read-only
+ push @$cmd, '-snapshot';
+ }
+
# add custom args
if ($conf->{args}) {
my $aa = PVE::Tools::split_args($conf->{args});
sub qemu_driveadd {
my ($storecfg, $vmid, $device) = @_;
- my $drive = print_drive_commandline_full($storecfg, $vmid, $device);
+ my $kvmver = get_running_qemu_version($vmid);
+ my $io_uring = min_version($kvmver, 6, 0);
+ my $drive = print_drive_commandline_full($storecfg, $vmid, $device, undef, $io_uring);
$drive =~ s/\\/\\\\/g;
my $ret = PVE::QemuServer::Monitor::hmp_cmd($vmid, "drive_add auto \"$drive\"");
my $padding = (1024 - $size % 1024) % 1024;
$size = $size + $padding;
- mon_cmd($vmid, "block_resize", device => $deviceid, size => int($size));
-
+ mon_cmd(
+ $vmid,
+ "block_resize",
+ device => $deviceid,
+ size => int($size),
+ timeout => 60,
+ );
}
sub qemu_volume_snapshot {
}
sub set_migration_caps {
- my ($vmid) = @_;
+ my ($vmid, $savevm) = @_;
my $qemu_support = eval { mon_cmd($vmid, "query-proxmox-support") };
+ my $bitmap_prop = $savevm ? 'pbs-dirty-bitmap-savevm' : 'pbs-dirty-bitmap-migration';
+ my $dirty_bitmaps = $qemu_support->{$bitmap_prop} ? 1 : 0;
+
my $cap_ref = [];
my $enabled_cap = {
"x-rdma-pin-all" => 0,
"zero-blocks" => 0,
"compress" => 0,
- "dirty-bitmaps" => $qemu_support->{'pbs-dirty-bitmap-migration'} ? 1 : 0,
+ "dirty-bitmaps" => $dirty_bitmaps,
};
my $supported_capabilities = mon_cmd($vmid, "query-migrate-capabilities");
$volhash->{$volid}->{is_unused} //= 0;
$volhash->{$volid}->{is_unused} = 1 if $key =~ /^unused\d+$/;
+
+ $volhash->{$volid}->{drivename} = $key if is_valid_drivename($key);
};
my $include_opts = {
die "skip\n" if !$hotplug_features->{memory};
PVE::QemuServer::Memory::qemu_memory_hotplug($vmid, $conf, $defaults, $opt);
} elsif ($opt eq 'cpuunits') {
- $cgroup->change_cpu_shares(undef, $defaults->{cpuunits});
+ $cgroup->change_cpu_shares(undef, 1024);
} elsif ($opt eq 'cpulimit') {
$cgroup->change_cpu_quota(-1, 100000);
} else {
die "skip\n" if !$hotplug_features->{memory};
$value = PVE::QemuServer::Memory::qemu_memory_hotplug($vmid, $conf, $defaults, $opt, $value);
} elsif ($opt eq 'cpuunits') {
- $cgroup->change_cpu_shares($conf->{pending}->{$opt}, $defaults->{cpuunits});
+ $cgroup->change_cpu_shares($conf->{pending}->{$opt}, 1024);
} elsif ($opt eq 'cpulimit') {
my $cpulimit = $conf->{pending}->{$opt} == 0 ? -1 : int($conf->{pending}->{$opt} * 100000);
$cgroup->change_cpu_quota($cpulimit, 100000);
# timeout => in seconds
# paused => start VM in paused state (backup)
# resume => resume from hibernation
+# pbs-backing => {
+# sata0 => {
+# repository
+# snapshot
+# keyfile
+# archive
+# },
+# virtio2 => ...
+# }
# migrate_opts:
# nbd => volumes for NBD exports (vm_migrate_alloc_nbd_disks)
# migratedfrom => source node
print "Resuming suspended VM\n";
}
- my ($cmd, $vollist, $spice_port) =
- config_to_command($storecfg, $vmid, $conf, $defaults, $forcemachine, $forcecpu);
+ my ($cmd, $vollist, $spice_port) = config_to_command($storecfg, $vmid,
+ $conf, $defaults, $forcemachine, $forcecpu, $params->{'pbs-backing'});
my $migration_ip;
my $get_migration_ip = sub {
# timeout should be more than enough here...
PVE::Systemd::wait_for_unit_removed("$vmid.scope", 5);
- my $cpuunits = defined($conf->{cpuunits}) ? $conf->{cpuunits}
- : $defaults->{cpuunits};
+ my $cpuunits = get_cpuunits($conf);
my $start_timeout = $params->{timeout} // config_aware_timeout($conf, $resume);
my %run_params = (
my %properties = (
Slice => 'qemu.slice',
- KillMode => 'none'
+ KillMode => 'process',
+ SendSIGKILL => 0,
+ TimeoutStopUSec => ULONG_MAX, # infinity
);
if (PVE::CGroup::cgroup_mode() == 2) {
+ $cpuunits = 10000 if $cpuunits >= 10000; # else we get an error
$properties{CPUWeight} = $cpuunits;
} else {
$properties{CPUShares} = $cpuunits;
PVE::Storage::activate_volumes($storecfg, [$vmstate]);
eval {
+ set_migration_caps($vmid, 1);
mon_cmd($vmid, "savevm-start", statefile => $path);
for(;;) {
my $state = mon_cmd($vmid, "query-savevm");
return $map;
};
-my $restore_update_config_line = sub {
- my ($outfd, $cookie, $vmid, $map, $line, $unique) = @_;
+sub restore_update_config_line {
+ my ($cookie, $map, $line, $unique) = @_;
- return if $line =~ m/^\#qmdump\#/;
- return if $line =~ m/^\#vzdump\#/;
- return if $line =~ m/^lock:/;
- return if $line =~ m/^unused\d+:/;
- return if $line =~ m/^parent:/;
+ return '' if $line =~ m/^\#qmdump\#/;
+ return '' if $line =~ m/^\#vzdump\#/;
+ return '' if $line =~ m/^lock:/;
+ return '' if $line =~ m/^unused\d+:/;
+ return '' if $line =~ m/^parent:/;
+
+ my $res = '';
my $dc = PVE::Cluster::cfs_read_file('datacenter.cfg');
if (($line =~ m/^(vlan(\d+)):\s*(\S+)\s*$/)) {
};
my $netstr = print_net($net);
- print $outfd "net$cookie->{netcount}: $netstr\n";
+ $res .= "net$cookie->{netcount}: $netstr\n";
$cookie->{netcount}++;
}
} elsif (($line =~ m/^(net\d+):\s*(\S+)\s*$/) && $unique) {
my $net = parse_net($netstr);
$net->{macaddr} = PVE::Tools::random_ether_addr($dc->{mac_prefix}) if $net->{macaddr};
$netstr = print_net($net);
- print $outfd "$id: $netstr\n";
+ $res .= "$id: $netstr\n";
} elsif ($line =~ m/^((ide|scsi|virtio|sata|efidisk)\d+):\s*(\S+)\s*$/) {
my $virtdev = $1;
my $value = $3;
my $di = parse_drive($virtdev, $value);
if (defined($di->{backup}) && !$di->{backup}) {
- print $outfd "#$line";
+ $res .= "#$line";
} elsif ($map->{$virtdev}) {
delete $di->{format}; # format can change on restore
$di->{file} = $map->{$virtdev};
$value = print_drive($di);
- print $outfd "$virtdev: $value\n";
+ $res .= "$virtdev: $value\n";
} else {
- print $outfd $line;
+ $res .= $line;
}
} elsif (($line =~ m/^vmgenid: (.*)/)) {
my $vmgenid = $1;
# always generate a new vmgenid if there was a valid one setup
$vmgenid = generate_uuid();
}
- print $outfd "vmgenid: $vmgenid\n";
+ $res .= "vmgenid: $vmgenid\n";
} elsif (($line =~ m/^(smbios1: )(.*)/) && $unique) {
my ($uuid, $uuid_str);
UUID::generate($uuid);
UUID::unparse($uuid, $uuid_str);
my $smbios1 = parse_smbios1($2);
$smbios1->{uuid} = $uuid_str;
- print $outfd $1.print_smbios1($smbios1)."\n";
+ $res .= $1.print_smbios1($smbios1)."\n";
} else {
- print $outfd $line;
+ $res .= $line;
}
-};
+
+ return $res;
+}
my $restore_deactivate_volumes = sub {
my ($storecfg, $devinfo) = @_;
sub scan_volids {
my ($cfg, $vmid) = @_;
- my $info = PVE::Storage::vdisk_list($cfg, undef, $vmid);
+ my $info = PVE::Storage::vdisk_list($cfg, undef, $vmid, undef, 'images');
my $volid_hash = {};
foreach my $storeid (keys %$info) {
my $cfg = PVE::Storage::config();
- # FIXME: Remove once our RBD plugin can handle CT and VM on a single storage
- # see: https://pve.proxmox.com/pipermail/pve-devel/2018-July/032900.html
- foreach my $stor (keys %{$cfg->{ids}}) {
- delete($cfg->{ids}->{$stor}) if ! $cfg->{ids}->{$stor}->{content}->{images};
- }
-
print "rescan volumes...\n";
my $volid_hash = scan_volids($cfg, $vmid);
my $repo = PVE::PBSClient::get_repository($scfg);
- # This is only used for `pbs-restore`!
+ # This is only used for `pbs-restore` and the QEMU PBS driver (live-restore)
my $password = PVE::Storage::PBSPlugin::pbs_get_password($scfg, $storeid);
local $ENV{PBS_PASSWORD} = $password;
local $ENV{PBS_FINGERPRINT} = $fingerprint if defined($fingerprint);
mkpath $tmpdir;
my $conffile = PVE::QemuConfig->config_file($vmid);
- my $tmpfn = "$conffile.$$.tmp";
# disable interrupts (always do cleanups)
local $SIG{INT} =
local $SIG{TERM} =
# Note: $oldconf is undef if VM does not exists
my $cfs_path = PVE::QemuConfig->cfs_config_path($vmid);
my $oldconf = PVE::Cluster::cfs_read_file($cfs_path);
+ my $new_conf_raw = '';
my $rpcenv = PVE::RPCEnvironment::get();
my $devinfo = {};
my $d = $virtdev_hash->{$virtdev};
next if $d->{is_cloudinit}; # no need to restore cloudinit
+ # this fails if storage is unavailable
my $volid = $d->{volid};
-
my $path = PVE::Storage::path($storecfg, $volid);
- # This is the ONLY user of the PBS_ env vars set on top of this function!
+ # for live-restore we only want to preload the efidisk
+ next if $options->{live} && $virtdev ne 'efidisk0';
+
my $pbs_restore_cmd = [
'/usr/bin/pbs-restore',
'--repository', $repo,
$fh->seek(0, 0) || die "seek failed - $!\n";
- my $outfd = IO::File->new($tmpfn, "w") || die "unable to write config for VM $vmid\n";
-
my $cookie = { netcount => 0 };
while (defined(my $line = <$fh>)) {
- $restore_update_config_line->($outfd, $cookie, $vmid, $map, $line, $options->{unique});
+ $new_conf_raw .= restore_update_config_line(
+ $cookie,
+ $map,
+ $line,
+ $options->{unique},
+ );
}
$fh->close();
- $outfd->close();
};
my $err = $@;
- $restore_deactivate_volumes->($storecfg, $devinfo);
+ if ($err || !$options->{live}) {
+ $restore_deactivate_volumes->($storecfg, $devinfo);
+ }
rmtree $tmpdir;
if ($err) {
- unlink $tmpfn;
$restore_destroy_volumes->($storecfg, $devinfo);
die $err;
}
- rename($tmpfn, $conffile) ||
- die "unable to commit configuration file '$conffile'\n";
+ if ($options->{live}) {
+ # keep lock during live-restore
+ $new_conf_raw .= "\nlock: create";
+ }
+
+ PVE::Tools::file_set_contents($conffile, $new_conf_raw);
PVE::Cluster::cfs_update(); # make sure we read new file
eval { rescan($vmid, 1); };
warn $@ if $@;
+
+ PVE::AccessControl::add_vm_to_pool($vmid, $options->{pool}) if $options->{pool};
+
+ if ($options->{live}) {
+ # enable interrupts
+ local $SIG{INT} =
+ local $SIG{TERM} =
+ local $SIG{QUIT} =
+ local $SIG{HUP} =
+ local $SIG{PIPE} = sub { die "got signal ($!) - abort\n"; };
+
+ my $conf = PVE::QemuConfig->load_config($vmid);
+ die "cannot do live-restore for template\n" if PVE::QemuConfig->is_template($conf);
+
+ delete $devinfo->{'drive-efidisk0'}; # this special drive is already restored before start
+ pbs_live_restore($vmid, $conf, $storecfg, $devinfo, $repo, $keyfile, $pbs_backup_name);
+
+ PVE::QemuConfig->remove_lock($vmid, "create");
+ }
+}
+
+sub pbs_live_restore {
+ my ($vmid, $conf, $storecfg, $restored_disks, $repo, $keyfile, $snap) = @_;
+
+ print "starting VM for live-restore\n";
+ print "repository: '$repo', snapshot: '$snap'\n";
+
+ my $pbs_backing = {};
+ for my $ds (keys %$restored_disks) {
+ $ds =~ m/^drive-(.*)$/;
+ my $confname = $1;
+ $pbs_backing->{$confname} = {
+ repository => $repo,
+ snapshot => $snap,
+ archive => "$ds.img.fidx",
+ };
+ $pbs_backing->{$confname}->{keyfile} = $keyfile if -e $keyfile;
+
+ my $drive = parse_drive($confname, $conf->{$confname});
+ print "restoring '$ds' to '$drive->{file}'\n";
+ }
+
+ my $drives_streamed = 0;
+ eval {
+ # make sure HA doesn't interrupt our restore by stopping the VM
+ if (PVE::HA::Config::vm_is_ha_managed($vmid)) {
+ run_command(['ha-manager', 'set', "vm:$vmid", '--state', 'started']);
+ }
+
+ # start VM with backing chain pointing to PBS backup, environment vars for PBS driver
+ # in QEMU (PBS_PASSWORD and PBS_FINGERPRINT) are already set by our caller
+ vm_start_nolock($storecfg, $vmid, $conf, {paused => 1, 'pbs-backing' => $pbs_backing}, {});
+
+ my $qmeventd_fd = register_qmeventd_handle($vmid);
+
+ # begin streaming, i.e. data copy from PBS to target disk for every vol,
+ # this will effectively collapse the backing image chain consisting of
+ # [target <- alloc-track -> PBS snapshot] to just [target] (alloc-track
+ # removes itself once all backing images vanish with 'auto-remove=on')
+ my $jobs = {};
+ for my $ds (sort keys %$restored_disks) {
+ my $job_id = "restore-$ds";
+ mon_cmd($vmid, 'block-stream',
+ 'job-id' => $job_id,
+ device => "$ds",
+ );
+ $jobs->{$job_id} = {};
+ }
+
+ mon_cmd($vmid, 'cont');
+ qemu_drive_mirror_monitor($vmid, undef, $jobs, 'auto', 0, 'stream');
+
+ print "restore-drive jobs finished successfully, removing all tracking block devices"
+ ." to disconnect from Proxmox Backup Server\n";
+
+ for my $ds (sort keys %$restored_disks) {
+ mon_cmd($vmid, 'blockdev-del', 'node-name' => "$ds-pbs");
+ }
+
+ close($qmeventd_fd);
+ };
+
+ my $err = $@;
+
+ if ($err) {
+ warn "An error occured during live-restore: $err\n";
+ _do_vm_stop($storecfg, $vmid, 1, 1, 10, 0, 1);
+ die "live-restore failed\n";
+ }
}
sub restore_vma_archive {
my $rpcenv = PVE::RPCEnvironment::get();
my $conffile = PVE::QemuConfig->config_file($vmid);
- my $tmpfn = "$conffile.$$.tmp";
# Note: $oldconf is undef if VM does not exist
my $cfs_path = PVE::QemuConfig->cfs_config_path($vmid);
my $oldconf = PVE::Cluster::cfs_read_file($cfs_path);
+ my $new_conf_raw = '';
my %storage_limits;
my $virtdev_hash = $parse_backup_hints->($rpcenv, $user, $cfg, $fh, $devinfo, $opts);
- foreach my $key (keys %storage_limits) {
- my $limit = PVE::Storage::get_bandwidth_limit('restore', [$key], $bwlimit);
- next if !$limit;
- print STDERR "rate limit for storage $key: $limit KiB/s\n";
- $storage_limits{$key} = $limit * 1024;
+ foreach my $info (values %{$virtdev_hash}) {
+ my $storeid = $info->{storeid};
+ next if defined($storage_limits{$storeid});
+
+ my $limit = PVE::Storage::get_bandwidth_limit('restore', [$storeid], $bwlimit) // 0;
+ print STDERR "rate limit for storage $storeid: $limit KiB/s\n" if $limit;
+ $storage_limits{$storeid} = $limit * 1024;
}
foreach my $devname (keys %$devinfo) {
$fh->seek(0, 0) || die "seek failed - $!\n";
- my $outfd = IO::File->new($tmpfn, "w") || die "unable to write config for VM $vmid\n";
-
my $cookie = { netcount => 0 };
while (defined(my $line = <$fh>)) {
- $restore_update_config_line->($outfd, $cookie, $vmid, $map, $line, $opts->{unique});
+ $new_conf_raw .= restore_update_config_line(
+ $cookie,
+ $map,
+ $line,
+ $opts->{unique},
+ );
}
$fh->close();
- $outfd->close();
};
eval {
rmtree $tmpdir;
if ($err) {
- unlink $tmpfn;
$restore_destroy_volumes->($cfg, $devinfo);
die $err;
}
- rename($tmpfn, $conffile) ||
- die "unable to commit configuration file '$conffile'\n";
+ PVE::Tools::file_set_contents($conffile, $new_conf_raw);
PVE::Cluster::cfs_update(); # make sure we read new file
eval { rescan($vmid, 1); };
warn $@ if $@;
+
+ PVE::AccessControl::add_vm_to_pool($vmid, $opts->{pool}) if $opts->{pool};
}
sub restore_tar_archive {
local $ENV{VZDUMP_USER} = $user;
my $conffile = PVE::QemuConfig->config_file($vmid);
- my $tmpfn = "$conffile.$$.tmp";
+ my $new_conf_raw = '';
# disable interrupts (always do cleanups)
local $SIG{INT} =
my $srcfd = IO::File->new($confsrc, "r") || die "unable to open file '$confsrc'\n";
- my $outfd = IO::File->new($tmpfn, "w") || die "unable to write config for VM $vmid\n";
-
my $cookie = { netcount => 0 };
while (defined (my $line = <$srcfd>)) {
- $restore_update_config_line->($outfd, $cookie, $vmid, $map, $line, $opts->{unique});
+ $new_conf_raw .= restore_update_config_line(
+ $cookie,
+ $map,
+ $line,
+ $opts->{unique},
+ );
}
$srcfd->close();
- $outfd->close();
};
if (my $err = $@) {
- unlink $tmpfn;
tar_restore_cleanup($storecfg, "$tmpdir/qmrestore.stat") if !$opts->{info};
die $err;
}
rmtree $tmpdir;
- rename $tmpfn, $conffile ||
- die "unable to commit configuration file '$conffile'\n";
+ PVE::Tools::file_set_contents($conffile, $new_conf_raw);
PVE::Cluster::cfs_update(); # make sure we read new file
if($line =~ m/\((\S+)\/100\%\)/){
my $percent = $1;
my $transferred = int($size * $percent / 100);
- my $remaining = $size - $transferred;
+ my $total_h = render_bytes($size, 1);
+ my $transferred_h = render_bytes($transferred, 1);
- print "transferred: $transferred bytes remaining: $remaining bytes total: $size bytes progression: $percent %\n";
+ print "transferred $transferred_h of $total_h ($percent%)\n";
}
};
# 'complete': wait until all jobs are ready, block-job-complete them (default)
# 'cancel': wait until all jobs are ready, block-job-cancel them
# 'skip': wait until all jobs are ready, return with block jobs in ready state
+# 'auto': wait until all jobs disappear, only use for jobs which complete automatically
sub qemu_drive_mirror_monitor {
- my ($vmid, $vmiddst, $jobs, $completion, $qga) = @_;
+ my ($vmid, $vmiddst, $jobs, $completion, $qga, $op) = @_;
$completion //= 'complete';
+ $op //= "mirror";
eval {
my $err_complete = 0;
+ my $starttime = time ();
while (1) {
- die "storage migration timed out\n" if $err_complete > 300;
+ die "block job ('$op') timed out\n" if $err_complete > 300;
my $stats = mon_cmd($vmid, "query-block-jobs");
+ my $ctime = time();
- my $running_mirror_jobs = {};
- foreach my $stat (@$stats) {
- next if $stat->{type} ne 'mirror';
- $running_mirror_jobs->{$stat->{device}} = $stat;
+ my $running_jobs = {};
+ for my $stat (@$stats) {
+ next if $stat->{type} ne $op;
+ $running_jobs->{$stat->{device}} = $stat;
}
my $readycounter = 0;
- foreach my $job (keys %$jobs) {
+ for my $job_id (sort keys %$jobs) {
+ my $job = $running_jobs->{$job_id};
- if(defined($jobs->{$job}->{complete}) && !defined($running_mirror_jobs->{$job})) {
- print "$job : finished\n";
- delete $jobs->{$job};
+ my $vanished = !defined($job);
+ my $complete = defined($jobs->{$job_id}->{complete}) && $vanished;
+ if($complete || ($vanished && $completion eq 'auto')) {
+ print "$job_id: $op-job finished\n";
+ delete $jobs->{$job_id};
next;
}
- die "$job: mirroring has been cancelled\n" if !defined($running_mirror_jobs->{$job});
+ die "$job_id: '$op' has been cancelled\n" if !defined($job);
- my $busy = $running_mirror_jobs->{$job}->{busy};
- my $ready = $running_mirror_jobs->{$job}->{ready};
- if (my $total = $running_mirror_jobs->{$job}->{len}) {
- my $transferred = $running_mirror_jobs->{$job}->{offset} || 0;
+ my $busy = $job->{busy};
+ my $ready = $job->{ready};
+ if (my $total = $job->{len}) {
+ my $transferred = $job->{offset} || 0;
my $remaining = $total - $transferred;
my $percent = sprintf "%.2f", ($transferred * 100 / $total);
- print "$job: transferred: $transferred bytes remaining: $remaining bytes total: $total bytes progression: $percent % busy: $busy ready: $ready \n";
+ my $duration = $ctime - $starttime;
+ my $total_h = render_bytes($total, 1);
+ my $transferred_h = render_bytes($transferred, 1);
+
+ my $status = sprintf(
+ "transferred $transferred_h of $total_h ($percent%%) in %s",
+ render_duration($duration),
+ );
+
+ if ($ready) {
+ if ($busy) {
+ $status .= ", still busy"; # shouldn't even happen? but mirror is weird
+ } else {
+ $status .= ", ready";
+ }
+ }
+ print "$job_id: $status\n" if !$jobs->{$job_id}->{ready};
+ $jobs->{$job_id}->{ready} = $ready;
}
- $readycounter++ if $running_mirror_jobs->{$job}->{ready};
+ $readycounter++ if $job->{ready};
}
last if scalar(keys %$jobs) == 0;
if ($readycounter == scalar(keys %$jobs)) {
- print "all mirroring jobs are ready \n";
- last if $completion eq 'skip'; #do the complete later
+ print "all '$op' jobs are ready\n";
+
+ # do the complete later (or has already been done)
+ last if $completion eq 'skip' || $completion eq 'auto';
if ($vmiddst && $vmiddst != $vmid) {
my $agent_running = $qga && qga_check_running($vmid);
last;
} else {
- foreach my $job (keys %$jobs) {
+ for my $job_id (sort keys %$jobs) {
# try to switch the disk if source and destination are on the same guest
- print "$job: Completing block job...\n";
+ print "$job_id: Completing block job_id...\n";
my $op;
if ($completion eq 'complete') {
} else {
die "invalid completion value: $completion\n";
}
- eval { mon_cmd($vmid, $op, device => $job) };
+ eval { mon_cmd($vmid, $op, device => $job_id) };
if ($@ =~ m/cannot be completed/) {
- print "$job: Block job cannot be completed, try again.\n";
+ print "$job_id: block job cannot be completed, trying again.\n";
$err_complete++;
}else {
- print "$job: Completed successfully.\n";
- $jobs->{$job}->{complete} = 1;
+ print "$job_id: Completed successfully.\n";
+ $jobs->{$job_id}->{complete} = 1;
}
}
}
if ($err) {
eval { PVE::QemuServer::qemu_blockjobs_cancel($vmid, $jobs) };
- die "mirroring error: $err";
+ die "block job ($op) error: $err";
}
-
}
sub qemu_blockjobs_cancel {
return $bootorder;
}
+sub register_qmeventd_handle {
+ my ($vmid) = @_;
+
+ my $fh;
+ my $peer = "/var/run/qmeventd.sock";
+ my $count = 0;
+
+ for (;;) {
+ $count++;
+ $fh = IO::Socket::UNIX->new(Peer => $peer, Blocking => 0, Timeout => 1);
+ last if $fh;
+ if ($! != EINTR && $! != EAGAIN) {
+ die "unable to connect to qmeventd socket (vmid: $vmid) - $!\n";
+ }
+ if ($count > 4) {
+ die "unable to connect to qmeventd socket (vmid: $vmid) - timeout "
+ . "after $count retries\n";
+ }
+ usleep(25000);
+ }
+
+ # send handshake to mark VM as backing up
+ print $fh to_json({vzdump => {vmid => "$vmid"}});
+
+ # return handle to be closed later when inhibit is no longer required
+ return $fh;
+}
+
# bash completion helper
sub complete_backup_archives {
return $qmpstatus && $qmpstatus->{status} eq "paused";
}
+sub check_volume_storage_type {
+ my ($storecfg, $vol) = @_;
+
+ my ($storeid, $volname) = PVE::Storage::parse_volume_id($vol);
+ my $scfg = PVE::Storage::storage_config($storecfg, $storeid);
+ my ($vtype) = PVE::Storage::parse_volname($storecfg, $vol);
+
+ die "storage '$storeid' does not support content-type '$vtype'\n"
+ if !$scfg->{content}->{$vtype};
+
+ return 1;
+}
+
1;
projects / qemu-server.git / blobdiff