X-Git-Url: https://git.proxmox.com/?a=blobdiff_plain;f=PVE%2FQemuMigrate.pm;h=cb1639bce70ac215254fbe056ca21f259113966b;hb=da18cc93001bfb70ec0b5f627389809c86effe8a;hp=2b383fa657041bea45e27bdf5ed8da7bdbfc9dd7;hpb=dabf24736c7eb7fd521f98c70940216f7f489208;p=qemu-server.git

diff --git a/PVE/QemuMigrate.pm b/PVE/QemuMigrate.pm
index 2b383fa..cb1639b 100644
--- a/PVE/QemuMigrate.pm
+++ b/PVE/QemuMigrate.pm
@@ -13,6 +13,9 @@ use PVE::Storage;
 use PVE::QemuServer;
 use Time::HiRes qw( usleep );
 use PVE::RPCEnvironment;
+use PVE::ReplicationConfig;
+use PVE::ReplicationState;
+use PVE::Replication;
 
 use base qw(PVE::AbstractMigrate);
 
@@ -93,7 +96,7 @@ sub fork_tunnel {
 
     my @localtunnelinfo = defined($tunnel_addr) ? ('-L' , $tunnel_addr ) : ();
 
-    my $cmd = [@{$self->{rem_ssh}}, '-o ExitOnForwardFailure=yes', @localtunnelinfo, 'qm', 'mtunnel' ];
+    my $cmd = [@{$self->{rem_ssh}}, '-o ExitOnForwardFailure=yes', @localtunnelinfo, '/usr/bin/pvecm', 'mtunnel' ];
 
     my $tunnel = $self->fork_command_pipe($cmd);
 
@@ -186,8 +189,10 @@ sub prepare {
 	my ($sid, $volname) = PVE::Storage::parse_volume_id($volid, 1);
 
 	# check if storage is available on both nodes
+	my $targetsid = $self->{opts}->{targetstorage} ? $self->{opts}->{targetstorage} : $sid;
+
 	my $scfg = PVE::Storage::storage_check_node($self->{storecfg}, $sid);
-	PVE::Storage::storage_check_node($self->{storecfg}, $sid, $self->{node});
+	PVE::Storage::storage_check_node($self->{storecfg}, $targetsid, $self->{node});
 
 	if ($scfg->{shared}) {
 	    # PVE::Storage::activate_storage checks this for non-shared storages
@@ -214,8 +219,6 @@ sub prepare {
 sub sync_disks {
     my ($self, $vmid) = @_;
 
-    $self->log('info', "copying disk images");
-
     my $conf = $self->{vmconf};
 
     # local volumes which have been copied
@@ -227,9 +230,23 @@ sub sync_disks {
 
 	# found local volumes and their origin
 	my $local_volumes = {};
+	my $local_volumes_errors = {};
+	my $other_errors = [];
+	my $abort = 0;
 
 	my $sharedvm = 1;
 
+	my $log_error = sub {
+	    my ($msg, $volid) = @_;
+
+	    if (defined($volid)) {
+		$local_volumes_errors->{$volid} = $msg;
+	    } else {
+		push @$other_errors, $msg;
+	    }
+	    $abort = 1;
+	};
+
 	my @sids = PVE::Storage::storage_ids($self->{storecfg});
 	foreach my $storeid (@sids) {
 	    my $scfg = PVE::Storage::storage_config($self->{storecfg}, $storeid);
@@ -241,90 +258,114 @@ sub sync_disks {
 
 	    next if @{$dl->{$storeid}} == 0;
 
+	    my $targetsid = $self->{opts}->{targetstorage} ? $self->{opts}->{targetstorage} : $storeid;
+
 	    # check if storage is available on target node
-	    PVE::Storage::storage_check_node($self->{storecfg}, $storeid, $self->{node});
+	    PVE::Storage::storage_check_node($self->{storecfg}, $targetsid, $self->{node});
 	    $sharedvm = 0; # there is a non-shared disk
 
 	    PVE::Storage::foreach_volid($dl, sub {
 		my ($volid, $sid, $volname) = @_;
 
-		$local_volumes->{$volid} = 'storage';
+		$local_volumes->{$volid}->{ref} = 'storage';
 	    });
 	}
 
 	my $test_volid = sub {
-	    my ($volid, $is_cdrom, $snapname) = @_;
+	    my ($volid, $attr) = @_;
 
-	    return if !$volid;
+	    if ($volid =~ m|^/|) {
+		$local_volumes->{$volid}->{ref} = 'config';
+		die "local file/device\n";
+	    }
 
-	    die "can't migrate local file/device '$volid'\n" if $volid =~ m|^/|;
+	    my $snaprefs = $attr->{referenced_in_snapshot};
 
-	    if ($is_cdrom) {
-		die "can't migrate local cdrom drive\n" if $volid eq 'cdrom';
+	    if ($attr->{cdrom}) {
+		if ($volid eq 'cdrom') {
+		    my $msg = "can't migrate local cdrom drive";
+		    if (defined($snaprefs) && !$attr->{referenced_in_config}) {
+			my $snapnames = join(', ', sort keys %$snaprefs);
+			$msg .= " (referenced in snapshot - $snapnames)";
+		    }
+		    &$log_error("$msg\n");
+		    return;
+		}
 		return if $volid eq 'none';
 	    }
 
 	    my ($sid, $volname) = PVE::Storage::parse_volume_id($volid);
 
+	    my $targetsid = $self->{opts}->{targetstorage} ? $self->{opts}->{targetstorage} : $sid;
 	    # check if storage is available on both nodes
 	    my $scfg = PVE::Storage::storage_check_node($self->{storecfg}, $sid);
-	    PVE::Storage::storage_check_node($self->{storecfg}, $sid, $self->{node});
+	    PVE::Storage::storage_check_node($self->{storecfg}, $targetsid, $self->{node});
 
 	    return if $scfg->{shared};
 
 	    $sharedvm = 0;
 
-	    $local_volumes->{$volid} = defined($snapname) ? 'snapshot' : 'config';
+	    $local_volumes->{$volid}->{ref} = $attr->{referenced_in_config} ? 'config' : 'snapshot';
 
-	    die "can't migrate local cdrom '$volid'\n" if $is_cdrom;
+	    die "local cdrom image\n" if $attr->{cdrom};
 
 	    my ($path, $owner) = PVE::Storage::path($self->{storecfg}, $volid);
 
-	    die "can't migrate volume '$volid' - owned by other VM (owner = VM $owner)\n"
+	    die "owned by other VM (owner = VM $owner)\n"
 		if !$owner || ($owner != $self->{vmid});
 
-	    if (defined($snapname)) {
+	    my $format = PVE::QemuServer::qemu_img_format($scfg, $volname);
+	    $local_volumes->{$volid}->{snapshots} = defined($snaprefs) || ($format =~ /^(?:qcow2|vmdk)$/);
+	    if (defined($snaprefs)) {
 		# we cannot migrate shapshots on local storage
 		# exceptions: 'zfspool' or 'qcow2' files (on directory storage)
 
-		my $format = PVE::QemuServer::qemu_img_format($scfg, $volname);
-
-		if (($scfg->{type} eq 'zfspool') || ($format eq 'qcow2')) {
-		    return;
+		die "online storage migration not possible if snapshot exists\n" if $self->{running};
+		if (!($scfg->{type} eq 'zfspool' || $format eq 'qcow2')) {
+		    die "non-migratable snapshot exists\n";
 		}
-
-		die "can't migrate snapshot of local volume '$volid'\n";
-
 	    }
-	};
 
-	my $test_drive = sub {
-	    my ($ds, $drive, $snapname) = @_;
-
-	    &$test_volid($drive->{file}, PVE::QemuServer::drive_is_cdrom($drive), $snapname);
+	    die "referenced by linked clone(s)\n"
+		if PVE::Storage::volume_is_base_and_used($self->{storecfg}, $volid);
 	};
 
-	foreach my $snapname (keys %{$conf->{snapshots}}) {
-	    &$test_volid($conf->{snapshots}->{$snapname}->{'vmstate'}, 0, undef)
-		if defined($conf->{snapshots}->{$snapname}->{'vmstate'});
-	    PVE::QemuServer::foreach_drive($conf->{snapshots}->{$snapname}, $test_drive, $snapname);
-	}
-	PVE::QemuServer::foreach_drive($conf, $test_drive);
+	PVE::QemuServer::foreach_volid($conf, sub {
+	    my ($volid, $attr) = @_;
+	    eval { $test_volid->($volid, $attr); };
+	    if (my $err = $@) {
+		&$log_error($err, $volid);
+	    }
+        });
 
 	foreach my $vol (sort keys %$local_volumes) {
-	    if ($local_volumes->{$vol} eq 'storage') {
+	    my $ref = $local_volumes->{$vol}->{ref};
+	    if ($ref eq 'storage') {
 		$self->log('info', "found local disk '$vol' (via storage)\n");
-	    } elsif ($local_volumes->{$vol} eq 'config') {
+	    } elsif ($ref eq 'config') {
+		&$log_error("can't live migrate attached local disks without with-local-disks option\n", $vol)
+		    if $self->{running} && !$self->{opts}->{"with-local-disks"};
 		$self->log('info', "found local disk '$vol' (in current VM config)\n");
-	    } elsif ($local_volumes->{$vol} eq 'snapshot') {
+	    } elsif ($ref eq 'snapshot') {
 		$self->log('info', "found local disk '$vol' (referenced by snapshot(s))\n");
 	    } else {
 		$self->log('info', "found local disk '$vol'\n");
 	    }
 	}
 
-	if ($self->{running} && !$sharedvm) {
-	    die "can't do online migration - VM uses local disks\n";
+	foreach my $vol (sort keys %$local_volumes_errors) {
+	    $self->log('warn', "can't migrate local disk '$vol': $local_volumes_errors->{$vol}");
+	}
+	foreach my $err (@$other_errors) {
+	    $self->log('warn', "$err");
+	}
+
+	if ($self->{running} && !$sharedvm && !$self->{opts}->{targetstorage}) {
+	    $self->{opts}->{targetstorage} = 1; #use same sid for remote local
+	}
+
+	if ($abort) {
+	    die "can't migrate VM - check log\n";
 	}
 
 	# additional checks for local storage
@@ -344,15 +385,55 @@ sub sync_disks {
 	    }
 	}
 
+	my $rep_volumes;
+
+	$self->log('info', "copying disk images");
+
+	my $rep_cfg = PVE::ReplicationConfig->new();
+
+	if (my $jobcfg = $rep_cfg->find_local_replication_job($vmid, $self->{node})) {
+	    die "can't live migrate VM with replicated volumes\n" if $self->{running};
+	    my $start_time = time();
+	    my $logfunc = sub { my ($msg) = @_;  $self->log('info', $msg); };
+	    $rep_volumes = PVE::Replication::run_replication(
+	       'PVE::QemuConfig', $jobcfg, $start_time, $start_time, $logfunc);
+	}
+
 	foreach my $volid (keys %$local_volumes) {
 	    my ($sid, $volname) = PVE::Storage::parse_volume_id($volid);
-	    push @{$self->{volumes}}, $volid;
-	    PVE::Storage::storage_migrate($self->{storecfg}, $volid, $self->{nodeip}, $sid);
+	    if ($self->{running} && $self->{opts}->{targetstorage} && $local_volumes->{$volid}->{ref} eq 'config') {
+		push @{$self->{online_local_volumes}}, $volid;
+	    } else {
+		next if $rep_volumes->{$volid};
+		push @{$self->{volumes}}, $volid;
+		my $insecure = $self->{opts}->{migration_type} eq 'insecure';
+		my $with_snapshots = $local_volumes->{$volid}->{snapshots};
+		PVE::Storage::storage_migrate($self->{storecfg}, $volid, $self->{ssh_info}, $sid,
+					      undef, undef, undef, undef, $insecure, $with_snapshots);
+	    }
 	}
     };
     die "Failed to sync data - $@" if $@;
 }
 
+sub cleanup_remotedisks {
+    my ($self) = @_;
+
+    foreach my $target_drive (keys %{$self->{target_drive}}) {
+
+	my $drive = PVE::QemuServer::parse_drive($target_drive, $self->{target_drive}->{$target_drive}->{volid});
+	my ($storeid, $volname) = PVE::Storage::parse_volume_id($drive->{file});
+
+	my $cmd = [@{$self->{rem_ssh}}, 'pvesm', 'free', "$storeid:$volname"];
+
+	eval{ PVE::Tools::run_command($cmd, outfunc => sub {}, errfunc => sub {}) };
+	if (my $err = $@) {
+	    $self->log('err', $err);
+	    $self->{errors} = 1;
+	}
+    }
+}
+
 sub phase1 {
     my ($self, $vmid) = @_;
 
@@ -411,14 +492,14 @@ sub phase2 {
 
     push @$cmd , 'qm', 'start', $vmid, '--skiplock', '--migratedfrom', $nodename;
 
-    # we use TCP only for unsecure migrations as TCP ssh forward tunnels often
-    # did appeared to late (they are hard, if not impossible, to check for)
-    # secure migration use UNIX sockets now, this *breaks* compatibilty when trying
-    # to migrate from new to old but *not* from old to new.
-    my $datacenterconf = PVE::Cluster::cfs_read_file('datacenter.cfg');
-    my $secure_migration = ($datacenterconf->{migration_unsecure}) ? 0 : 1;
+    my $migration_type = $self->{opts}->{migration_type};
 
-    if (!$secure_migration) {
+    push @$cmd, '--migration_type', $migration_type;
+
+    push @$cmd, '--migration_network', $self->{opts}->{migration_network}
+      if $self->{opts}->{migration_network};
+
+    if ($migration_type eq 'insecure') {
 	push @$cmd, '--stateuri', 'tcp';
     } else {
 	push @$cmd, '--stateuri', 'unix';
@@ -428,6 +509,10 @@ sub phase2 {
 	push @$cmd, '--machine', $self->{forcemachine};
     }
 
+    if ($self->{opts}->{targetstorage}) {
+	push @$cmd, '--targetstorage', $self->{opts}->{targetstorage};
+    }
+
     my $spice_port;
 
     # Note: We try to keep $spice_ticket secret (do not pass via command line parameter)
@@ -453,6 +538,16 @@ sub phase2 {
         elsif ($line =~ m/^spice listens on port (\d+)$/) {
 	    $spice_port = int($1);
 	}
+        elsif ($line =~ m/^storage migration listens on nbd:(localhost|[\d\.]+|\[[\d\.:a-fA-F]+\]):(\d+):exportname=(\S+) volume:(\S+)$/) {
+	    my $volid = $4;
+	    my $nbd_uri = "nbd:$1:$2:exportname=$3";
+	    my $targetdrive = $3;
+	    $targetdrive =~ s/drive-//g;
+
+	    $self->{target_drive}->{$targetdrive}->{volid} = $volid;
+	    $self->{target_drive}->{$targetdrive}->{nbd_uri} = $nbd_uri;
+
+	}
     }, errfunc => sub {
 	my $line = shift;
 	$self->log('info', $line);
@@ -460,7 +555,7 @@ sub phase2 {
 
     die "unable to detect remote migration address\n" if !$raddr;
 
-    if ($secure_migration) {
+    if ($migration_type eq 'secure') {
 	$self->log('info', "start remote tunnel");
 
 	if ($ruri =~ /^unix:/) {
@@ -497,6 +592,21 @@ sub phase2 {
     }
 
     my $start = time();
+
+    if ($self->{opts}->{targetstorage} && defined($self->{online_local_volumes})) {
+	$self->{storage_migration} = 1;
+	$self->{storage_migration_jobs} = {};
+	$self->log('info', "starting storage migration");
+
+	die "The number of local disks does not match between the source and the destination.\n"
+	    if (scalar(keys %{$self->{target_drive}}) != scalar @{$self->{online_local_volumes}});
+	foreach my $drive (keys %{$self->{target_drive}}){
+	    my $nbd_uri = $self->{target_drive}->{$drive}->{nbd_uri};
+	    $self->log('info', "$drive: start migration to to $nbd_uri");
+	    PVE::QemuServer::qemu_drive_mirror($vmid, $drive, $nbd_uri, $vmid, undef, $self->{storage_migration_jobs}, 1);
+	}
+    }
+
     $self->log('info', "starting online/live migration on $ruri");
     $self->{livemigration} = 1;
 
@@ -696,6 +806,19 @@ sub phase2_cleanup {
     }
 
     # cleanup ressources on target host
+    if ($self->{storage_migration}) {
+
+	eval { PVE::QemuServer::qemu_blockjobs_cancel($vmid, $self->{storage_migration_jobs}) };
+	if (my $err = $@) {
+	    $self->log('err', $err);
+	}
+
+	eval { PVE::QemuMigrate::cleanup_remotedisks($self) };
+	if (my $err = $@) {
+	    $self->log('err', $err);
+	}
+    }
+
     my $nodename = PVE::INotify::nodename();
  
     my $cmd = [@{$self->{rem_ssh}}, 'qm', 'stop', $vmid, '--skiplock', '--migratedfrom', $nodename];
@@ -737,6 +860,26 @@ sub phase3_cleanup {
     my $conf = $self->{vmconf};
     return if $self->{phase2errors};
 
+    if ($self->{storage_migration}) {
+	# finish block-job
+	eval { PVE::QemuServer::qemu_drive_mirror_monitor($vmid, undef, $self->{storage_migration_jobs}); };
+
+	if (my $err = $@) {
+	    eval { PVE::QemuServer::qemu_blockjobs_cancel($vmid, $self->{storage_migration_jobs}) };
+	    eval { PVE::QemuMigrate::cleanup_remotedisks($self) };
+	    die "Failed to completed storage migration\n";
+	} else {
+	    foreach my $target_drive (keys %{$self->{target_drive}}) {
+		my $drive = PVE::QemuServer::parse_drive($target_drive, $self->{target_drive}->{$target_drive}->{volid});
+		$conf->{$target_drive} = PVE::QemuServer::print_drive($vmid, $drive);
+		PVE::QemuConfig->write_config($vmid, $conf);
+	    }
+	}
+    }
+
+    # transfer replication state before move config
+    $self->transfer_replication_state();
+
     # move config to remote node
     my $conffile = PVE::QemuConfig->config_file($vmid);
     my $newconffile = PVE::QemuConfig->config_file($vmid, $self->{node});
@@ -744,8 +887,25 @@ sub phase3_cleanup {
     die "Failed to move config to node '$self->{node}' - rename failed: $!\n"
         if !rename($conffile, $newconffile);
 
+    $self->switch_replication_job_target();
+
     if ($self->{livemigration}) {
-	# now that config file is move, we can resume vm on target if livemigrate
+	if ($self->{storage_migration}) {
+	    # remove drives referencing the nbd server from source
+	    # otherwise vm_stop might hang later on
+	    foreach my $drive (keys %{$self->{target_drive}}){
+		PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "device_del", id => $drive);
+	    }
+	    # stop nbd server on remote vm - requirement for resume since 2.9
+	    my $cmd = [@{$self->{rem_ssh}}, 'qm', 'nbdstop', $vmid];
+
+	    eval{ PVE::Tools::run_command($cmd, outfunc => sub {}, errfunc => sub {}) };
+	    if (my $err = $@) {
+		$self->log('err', $err);
+		$self->{errors} = 1;
+	    }
+	}
+	# config moved and nbd server stopped - now we can resume vm on target
 	my $cmd = [@{$self->{rem_ssh}}, 'qm', 'resume', $vmid, '--skiplock', '--nocheck'];
 	eval{ PVE::Tools::run_command($cmd, outfunc => sub {}, 
 		errfunc => sub {
@@ -760,7 +920,6 @@ sub phase3_cleanup {
     }
 
     eval {
-
 	my $timer = 0;
 	if (PVE::QemuServer::vga_conf_has_spice($conf->{vga}) && $self->{running}) {
 	    $self->log('info', "Waiting for spice server migration");
@@ -791,6 +950,21 @@ sub phase3_cleanup {
 	$self->{errors} = 1;
     }
 
+    if($self->{storage_migration}) {
+	# destroy local copies
+	my $volids = $self->{online_local_volumes};
+
+	foreach my $volid (@$volids) {
+	    eval { PVE::Storage::vdisk_free($self->{storecfg}, $volid); };
+	    if (my $err = $@) {
+		$self->log('err', "removing local copy of '$volid' failed - $err");
+		$self->{errors} = 1;
+		last if $err =~ /^interrupted by signal$/;
+	    }
+	}
+
+    }
+
     # clear migrate lock
     my $cmd = [ @{$self->{rem_ssh}}, 'qm', 'unlock', $vmid ];
     $self->cmd_logerr($cmd, errmsg => "failed to clear migrate lock");