optional => 1,
});
-PVE::JSONSchema::register_standard_option('pve-qemu-machine', {
- description => "Specifies the QEMU machine type.",
- type => 'string',
- pattern => '(pc|pc(-i440fx)?-\d+(\.\d+)+(\+pve\d+)?(\.pxe)?|q35|pc-q35-\d+(\.\d+)+(\+pve\d+)?(\.pxe)?|virt(?:-\d+(\.\d+)+)?(\+pve\d+)?)',
- maxLength => 40,
- optional => 1,
-});
-
# FIXME: remove in favor of just using the INotify one, it's cached there exactly the same way
my $nodename_cache;
sub nodename {
ostype => {
optional => 1,
type => 'string',
+ # NOTE: When extending, also consider extending `%guest_types` in `Import/ESXi.pm`.
enum => [qw(other wxp w2k w2k3 w2k8 wvista win7 win8 win10 win11 l24 l26 solaris)],
description => "Specify guest operating system.",
verbose_description => <<EODESC,
win7;; Microsoft Windows 7
win8;; Microsoft Windows 8/2012/2012r2
win10;; Microsoft Windows 10/2016/2019
-win11;; Microsoft Windows 11/2022
+win11;; Microsoft Windows 11/2022/2025
l24;; Linux 2.4 Kernel
l26;; Linux 2.6 - 6.X Kernel
solaris;; Solaris/OpenSolaris/OpenIndiania kernel
migrate_downtime => {
optional => 1,
type => 'number',
- description => "Set maximum tolerated downtime (in seconds) for migrations.",
+ description => "Set maximum tolerated downtime (in seconds) for migrations. Should the"
+ ." migration not be able to converge in the very end, because too much newly dirtied"
+ ." RAM needs to be transferred, the limit will be increased automatically step-by-step"
+ ." until migration can converge.",
minimum => 0,
default => 0.1,
},
my $unit = $drive->{index} % $maxdev;
my $machine_version = extract_version($machine_type, kvm_user_version());
- my $devicetype = PVE::QemuServer::Drive::get_scsi_devicetype(
+ my $device_type = PVE::QemuServer::Drive::get_scsi_device_type(
$drive, $storecfg, $machine_version);
if (!$conf->{scsihw} || $conf->{scsihw} =~ m/^lsi/ || $conf->{scsihw} eq 'pvscsi') {
- $device = "scsi-$devicetype,bus=$controller_prefix$controller.0,scsi-id=$unit";
+ $device = "scsi-$device_type,bus=$controller_prefix$controller.0,scsi-id=$unit";
} else {
- $device = "scsi-$devicetype,bus=$controller_prefix$controller.0,channel=0,scsi-id=0"
+ $device = "scsi-$device_type,bus=$controller_prefix$controller.0,channel=0,scsi-id=0"
.",lun=$drive->{index}";
}
$device .= ",drive=drive-$drive_id,id=$drive_id";
- if ($drive->{ssd} && ($devicetype eq 'block' || $devicetype eq 'hd')) {
+ if ($drive->{ssd} && ($device_type eq 'block' || $device_type eq 'hd')) {
$device .= ",rotation_rate=1";
}
$device .= ",wwn=$drive->{wwn}" if $drive->{wwn};
# only scsi-hd and scsi-cd support passing vendor and product information
- if ($devicetype eq 'hd' || $devicetype eq 'cd') {
+ if ($device_type eq 'hd' || $device_type eq 'cd') {
if (my $vendor = $drive->{vendor}) {
$device .= ",vendor=$vendor";
}
$unit = 0;
}
- my $devicetype = ($drive->{media} && $drive->{media} eq 'cdrom') ? "cd" : "hd";
+ my $device_type = ($drive->{media} && $drive->{media} eq 'cdrom') ? "cd" : "hd";
- $device = "ide-$devicetype";
+ $device = "ide-$device_type";
if ($drive->{interface} eq 'ide') {
$device .= ",bus=ide.$controller,unit=$unit";
} else {
}
$device .= ",drive=drive-$drive_id,id=$drive_id";
- if ($devicetype eq 'hd') {
+ if ($device_type eq 'hd') {
if (my $model = $drive->{model}) {
$model = URI::Escape::uri_unescape($model);
$device .= ",model=$model";
}
sub print_drive_commandline_full {
- my ($storecfg, $vmid, $drive, $pbs_name, $io_uring) = @_;
+ my ($storecfg, $vmid, $drive, $live_restore_name, $io_uring) = @_;
my $path;
my $volid = $drive->{file};
if (drive_is_cdrom($drive)) {
$path = get_iso_path($storecfg, $vmid, $volid);
- die "$drive_id: cannot back cdrom drive with PBS snapshot\n" if $pbs_name;
+ die "$drive_id: cannot back cdrom drive with a live restore image\n" if $live_restore_name;
} else {
if ($storeid) {
$path = PVE::Storage::path($storecfg, $volid);
}
}
- if ($pbs_name) {
+ if ($live_restore_name) {
$format = "rbd" if $is_rbd;
die "$drive_id: Proxmox Backup Server backed drive cannot auto-detect the format\n"
if !$format;
# 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";
+ my $dz_param = $live_restore_name ? "file.detect-zeroes" : "detect-zeroes";
$opts .= ",$dz_param=$detectzeroes" if $detectzeroes;
}
- if ($pbs_name) {
- $opts .= ",backing=$pbs_name";
+ if ($live_restore_name) {
+ $opts .= ",backing=$live_restore_name";
$opts .= ",auto-remove=on";
}
- # my $file_param = $pbs_name ? "file.file.filename" : "file";
+ # my $file_param = $live_restore_name ? "file.file.filename" : "file";
my $file_param = "file";
- if ($pbs_name) {
+ if ($live_restore_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;
}
}
- die "no devicetype for $vga->{type}\n" if !$type;
+ die "no device-type for $vga->{type}\n" if !$type;
my $memory = "";
if ($vgamem_mb) {
# check if we need to apply some handling for VMs that always use the latest machine version but
# had a machine version transition happen that affected HW such that, e.g., an OS config change
# would be required (we do not want to pin machine version for non-windows OS type)
+ my $machine_conf = PVE::QemuServer::Machine::parse_machine($conf->{machine});
if (
- (!defined($conf->{machine}) || $conf->{machine} =~ m/^(?:pc|q35|virt)$/) # non-versioned machine
+ (!defined($machine_conf->{type}) || $machine_conf->{type} =~ m/^(?:pc|q35|virt)$/) # non-versioned machine
&& (!defined($meta->{'creation-qemu'}) || !min_version($meta->{'creation-qemu'}, 6, 1)) # created before 6.1
&& (!$forced_vers || min_version($forced_vers, 6, 1)) # handle snapshot-rollback/migrations
&& min_version($kvmver, 6, 1) # only need to apply the change since 6.1
foreach my $k (keys %$conf) {
if ($k =~ m/^usb/) {
my $entry = parse_property_string('pve-qm-usb', $conf->{$k});
- next if $entry->{host} =~ m/^spice$/i;
+ next if $entry->{host} && $entry->{host} =~ m/^spice$/i;
if ($entry->{mapping}) {
$add_missing_mapping->('usb', $k, $entry->{mapping});
push @$mapped_res, $k;
sub get_vm_machine {
my ($conf, $forcemachine, $arch, $add_pve_version, $kvmversion) = @_;
- my $machine = $forcemachine || $conf->{machine};
+ my $machine_conf = PVE::QemuServer::Machine::parse_machine($conf->{machine});
+ my $machine = $forcemachine || $machine_conf->{type};
if (!$machine || $machine =~ m/^(?:pc|q35|virt)$/) {
$kvmversion //= kvm_user_version();
sub config_to_command {
my ($storecfg, $vmid, $conf, $defaults, $forcemachine, $forcecpu,
- $pbs_backing) = @_;
+ $live_restore_backing) = @_;
my ($globalFlags, $machineFlags, $rtcFlags) = ([], [], []);
my $devices = [];
my $kvm = $conf->{kvm};
my $nodename = nodename();
+ my $machine_conf = PVE::QemuServer::Machine::parse_machine($conf->{machine});
+
my $arch = get_vm_arch($conf);
my $kvm_binary = get_command_for_arch($arch);
my $kvmver = kvm_user_version($kvm_binary);
$ahcicontroller->{$controller}=1;
}
- 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 $live_restore = $live_restore_backing->{$ds};
+ my $live_blockdev_name = undef;
+ if ($live_restore) {
+ $live_blockdev_name = $live_restore->{name};
+ push @$devices, '-blockdev', $live_restore->{blockdev};
}
my $drive_cmd = print_drive_commandline_full(
- $storecfg, $vmid, $drive, $pbs_name, min_version($kvmver, 6, 0));
+ $storecfg, $vmid, $drive, $live_blockdev_name, min_version($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, $drive);
}
push @$machineFlags, "type=${machine_type_min}";
+ PVE::QemuServer::Machine::assert_valid_machine_property($conf, $machine_conf);
+
+ if (my $viommu = $machine_conf->{viommu}) {
+ if ($viommu eq 'intel') {
+ unshift @$devices, '-device', 'intel-iommu,intremap=on,caching-mode=on';
+ push @$machineFlags, 'kernel-irqchip=split';
+ } elsif ($viommu eq 'virtio') {
+ push @$devices, '-device', 'virtio-iommu-pci';
+ }
+ }
+
push @$cmd, @$devices;
push @$cmd, '-rtc', join(',', @$rtcFlags) if scalar(@$rtcFlags);
push @$cmd, '-machine', join(',', @$machineFlags) if scalar(@$machineFlags);
if (defined($conf->{$opt}) && is_valid_drivename($opt)) {
vmconfig_register_unused_drive($storecfg, $vmid, $conf, parse_drive($opt, $conf->{$opt}))
} elsif (defined($conf->{pending}->{$opt}) && $opt =~ m/^net\d+$/) {
- if($have_sdn) {
- my $new_net = PVE::QemuServer::parse_net($conf->{pending}->{$opt});
- if ($conf->{$opt}){
- my $old_net = PVE::QemuServer::parse_net($conf->{$opt});
-
- if (defined($old_net->{bridge}) && defined($old_net->{macaddr}) && (
- safe_string_ne($old_net->{bridge}, $new_net->{bridge}) ||
- safe_string_ne($old_net->{macaddr}, $new_net->{macaddr})
- )) {
- PVE::Network::SDN::Vnets::del_ips_from_mac($old_net->{bridge}, $old_net->{macaddr}, $conf->{name});
- }
- }
- #fixme: reuse ip if mac change && same bridge
- PVE::Network::SDN::Vnets::add_next_free_cidr($new_net->{bridge}, $conf->{name}, $new_net->{macaddr}, $vmid, undef, 1);
+ return if !$have_sdn; # return from eval if SDN is not available
+
+ my $new_net = PVE::QemuServer::parse_net($conf->{pending}->{$opt});
+ if ($conf->{$opt}) {
+ my $old_net = PVE::QemuServer::parse_net($conf->{$opt});
+
+ if (defined($old_net->{bridge}) && defined($old_net->{macaddr}) && (
+ safe_string_ne($old_net->{bridge}, $new_net->{bridge}) ||
+ safe_string_ne($old_net->{macaddr}, $new_net->{macaddr})
+ )) {
+ PVE::Network::SDN::Vnets::del_ips_from_mac($old_net->{bridge}, $old_net->{macaddr}, $conf->{name});
+ }
}
+ #fixme: reuse ip if mac change && same bridge
+ PVE::Network::SDN::Vnets::add_next_free_cidr($new_net->{bridge}, $conf->{name}, $new_net->{macaddr}, $vmid, undef, 1);
}
};
if (my $err = $@) {
safe_string_ne($oldnet->{macaddr}, $newnet->{macaddr}) ||
safe_num_ne($oldnet->{queues}, $newnet->{queues}) ||
safe_num_ne($oldnet->{mtu}, $newnet->{mtu}) ||
- !($newnet->{bridge} && $oldnet->{bridge})) { # bridge/nat mode change
+ !($newnet->{bridge} && $oldnet->{bridge})
+ ) { # bridge/nat mode change
# for non online change, we try to hot-unplug
die "skip\n" if !$hotplug;
vm_deviceunplug($vmid, $conf, $opt);
- if($have_sdn) {
+ if ($have_sdn) {
PVE::Network::SDN::Vnets::del_ips_from_mac($oldnet->{bridge}, $oldnet->{macaddr}, $conf->{name});
}
if (safe_string_ne($oldnet->{bridge}, $newnet->{bridge}) ||
safe_num_ne($oldnet->{tag}, $newnet->{tag}) ||
safe_string_ne($oldnet->{trunks}, $newnet->{trunks}) ||
- safe_num_ne($oldnet->{firewall}, $newnet->{firewall})) {
+ safe_num_ne($oldnet->{firewall}, $newnet->{firewall})
+ ) {
PVE::Network::tap_unplug($iface);
#set link_down in guest if bridge or vlan change to notify guest (dhcp renew for example)
if (safe_string_ne($oldnet->{bridge}, $newnet->{bridge}) ||
- safe_num_ne($oldnet->{tag}, $newnet->{tag})) {
+ safe_num_ne($oldnet->{tag}, $newnet->{tag})
+ ) {
qemu_set_link_status($vmid, $opt, 0);
}
#set link_up in guest if bridge or vlan change to notify guest (dhcp renew for example)
if (safe_string_ne($oldnet->{bridge}, $newnet->{bridge}) ||
- safe_num_ne($oldnet->{tag}, $newnet->{tag})) {
+ safe_num_ne($oldnet->{tag}, $newnet->{tag})
+ ) {
qemu_set_link_status($vmid, $opt, 1);
}
# timeout => in seconds
# paused => start VM in paused state (backup)
# resume => resume from hibernation
-# pbs-backing => {
+# live-restore-backing => {
# sata0 => {
-# repository
-# snapshot
-# keyfile
-# archive
+# name => blockdev-name,
+# blockdev => "arg to the -blockdev command instantiating device named 'name'",
# },
# virtio2 => ...
# }
}
my ($cmd, $vollist, $spice_port, $pci_devices) = config_to_command($storecfg, $vmid,
- $conf, $defaults, $forcemachine, $forcecpu, $params->{'pbs-backing'});
+ $conf, $defaults, $forcemachine, $forcecpu, $params->{'live-restore-backing'});
my $migration_ip;
my $get_migration_ip = sub {
if ($err) {
# cleanup, but leave suspending lock, to indicate something went wrong
eval {
- mon_cmd($vmid, "savevm-end");
+ eval { mon_cmd($vmid, "savevm-end"); };
+ warn $@ if $@;
PVE::Storage::deactivate_volumes($storecfg, [$vmstate]);
PVE::Storage::vdisk_free($storecfg, $vmstate);
delete $conf->@{qw(vmstate runningmachine runningcpu)};
print "starting VM for live-restore\n";
print "repository: '$opts->{repo}', snapshot: '$opts->{snapshot}'\n";
- my $pbs_backing = {};
+ my $live_restore_backing = {};
for my $ds (keys %$restored_disks) {
$ds =~ m/^drive-(.*)$/;
my $confname = $1;
- $pbs_backing->{$confname} = {
+ my $pbs_conf = {};
+ $pbs_conf = {
repository => $opts->{repo},
snapshot => $opts->{snapshot},
archive => "$ds.img.fidx",
};
- $pbs_backing->{$confname}->{keyfile} = $opts->{keyfile} if -e $opts->{keyfile};
- $pbs_backing->{$confname}->{namespace} = $opts->{namespace} if defined($opts->{namespace});
+ $pbs_conf->{keyfile} = $opts->{keyfile} if -e $opts->{keyfile};
+ $pbs_conf->{namespace} = $opts->{namespace} if defined($opts->{namespace});
my $drive = parse_drive($confname, $conf->{$confname});
print "restoring '$ds' to '$drive->{file}'\n";
+
+ my $pbs_name = "drive-${confname}-pbs";
+ $live_restore_backing->{$confname} = {
+ name => $pbs_name,
+ blockdev => print_pbs_blockdev($pbs_conf, $pbs_name),
+ };
}
my $drives_streamed = 0;
# start VM with backing chain pointing to PBS backup, environment vars for PBS driver
# in QEMU (PBS_PASSWORD and PBS_FINGERPRINT) are already set by our caller
- vm_start_nolock($storecfg, $vmid, $conf, {paused => 1, 'pbs-backing' => $pbs_backing}, {});
+ vm_start_nolock($storecfg, $vmid, $conf, {paused => 1, 'live-restore-backing' => $live_restore_backing}, {});
my $qmeventd_fd = register_qmeventd_handle($vmid);
}
}
+# Inspired by pbs live-restore, this restores with the disks being available as files.
+# Theoretically this can also be used to quick-start a full-clone vm if the
+# disks are all available as files.
+#
+# The mapping should provide a path by config entry, such as
+# `{ scsi0 => { format => <qcow2|raw|...>, path => "/path/to/file", sata1 => ... } }`
+#
+# This is used when doing a `create` call with the `--live-import` parameter,
+# where the disks get an `import-from=` property. The non-live part is
+# therefore already handled in the `$create_disks()` call happening in the
+# `create` api call
+sub live_import_from_files {
+ my ($mapping, $vmid, $conf, $restore_options) = @_;
+
+ my $live_restore_backing = {};
+ for my $dev (keys %$mapping) {
+ die "disk not support for live-restoring: '$dev'\n"
+ if !is_valid_drivename($dev) || $dev =~ /^(?:efidisk|tpmstate)/;
+
+ die "mapping contains disk '$dev' which does not exist in the config\n"
+ if !exists($conf->{$dev});
+
+ my $info = $mapping->{$dev};
+ my ($format, $path) = $info->@{qw(format path)};
+ die "missing path for '$dev' mapping\n" if !$path;
+ die "missing format for '$dev' mapping\n" if !$format;
+ die "invalid format '$format' for '$dev' mapping\n"
+ if !grep { $format eq $_ } qw(raw qcow2 vmdk);
+
+ $live_restore_backing->{$dev} = {
+ name => "drive-$dev-restore",
+ blockdev => "driver=$format,node-name=drive-$dev-restore"
+ . ",read-only=on"
+ . ",file.driver=file,file.filename=$path"
+ };
+ };
+
+ my $storecfg = PVE::Storage::config();
+ eval {
+
+ # make sure HA doesn't interrupt our restore by stopping the VM
+ if (PVE::HA::Config::vm_is_ha_managed($vmid)) {
+ run_command(['ha-manager', 'set', "vm:$vmid", '--state', 'started']);
+ }
+
+ vm_start_nolock($storecfg, $vmid, $conf, {paused => 1, 'live-restore-backing' => $live_restore_backing}, {});
+
+ # prevent shutdowns from qmeventd when the VM powers off from the inside
+ 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 $ds (sort keys %$live_restore_backing) {
+ my $job_id = "restore-$ds";
+ mon_cmd($vmid, 'block-stream',
+ 'job-id' => $job_id,
+ device => "drive-$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\n";
+
+ for my $ds (sort keys %$live_restore_backing) {
+ mon_cmd($vmid, 'blockdev-del', 'node-name' => "drive-$ds-restore");
+ }
+
+ close($qmeventd_fd);
+ };
+
+ my $err = $@;
+
+ if ($err) {
+ warn "An error occurred during live-restore: $err\n";
+ _do_vm_stop($storecfg, $vmid, 1, 1, 10, 0, 1);
+ die "live-restore failed\n";
+ }
+
+ PVE::QemuConfig->remove_lock($vmid, "import");
+}
+
sub restore_vma_archive {
my ($archive, $vmid, $user, $opts, $comp) = @_;
sub qemu_img_format {
my ($scfg, $volname) = @_;
- if ($scfg->{path} && $volname =~ m/\.($PVE::QemuServer::Drive::QEMU_FORMAT_RE)$/) {
+ # FIXME: this entire function is kind of weird given that `parse_volname`
+ # also already gives us a format?
+ my $is_path_storage = $scfg->{path} || $scfg->{type} eq 'esxi';
+
+ if ($is_path_storage && $volname =~ m/\.($PVE::QemuServer::Drive::QEMU_FORMAT_RE)$/) {
return $1;
} else {
return "raw";
my ($newvmid, $dst_drivename, $efisize) = $dest->@{qw(vmid drivename efisize)};
my ($storage, $format) = $dest->@{qw(storage format)};
- my $use_drive_mirror = $full && $running && $src_drivename && !$snapname;
+ my $unused = defined($src_drivename) && $src_drivename =~ /^unused/;
+ my $use_drive_mirror = $full && $running && $src_drivename && !$snapname && !$unused;
if ($src_drivename && $dst_drivename && $src_drivename ne $dst_drivename) {
die "cloning from/to EFI disk requires EFI disk\n"
my $disk = dclone($drive);
delete $disk->{format};
$disk->{file} = $newvolid;
- $disk->{size} = $size if defined($size);
+ $disk->{size} = $size if defined($size) && !$unused;
return $disk;
}
projects / qemu-server.git / blobdiff