[pve-devel] [PATCH qemu-server 5/5] phase2: add livemigrate
Alexandre Derumier
aderumier at odiso.com
Thu Apr 25 09:22:53 CEST 2019
Signed-off-by: Alexandre Derumier <aderumier at odiso.com>
---
PVE/QemuMigrate.pm | 370 +++++++++++++++++++++++++++--------------------------
1 file changed, 187 insertions(+), 183 deletions(-)
diff --git a/PVE/QemuMigrate.pm b/PVE/QemuMigrate.pm
index b38bf5c..2bca615 100644
--- a/PVE/QemuMigrate.pm
+++ b/PVE/QemuMigrate.pm
@@ -538,8 +538,6 @@ sub phase2 {
my $nodename = PVE::INotify::nodename();
- my $conf = $self->{vmconf};
-
$self->log('info', "starting VM $vmid on remote node '$self->{node}'");
my $migration_type = $self->{opts}->{migration_type};
@@ -552,188 +550,8 @@ sub phase2 {
livemigrate_storage($self, $vmid);
- my $start = time();
-
- my $opt_bwlimit = $self->{opts}->{bwlimit};
-
- $self->log('info', "starting online/live migration on $ruri");
- $self->{livemigration} = 1;
-
- # load_defaults
- my $defaults = PVE::QemuServer::load_defaults();
-
- # migrate speed can be set via bwlimit (datacenter.cfg and API) and via the
- # migrate_speed parameter in qm.conf - take the lower of the two.
- my $bwlimit = PVE::Storage::get_bandwidth_limit('migrate', undef, $opt_bwlimit) // 0;
- my $migrate_speed = $conf->{migrate_speed} // $bwlimit;
- # migrate_speed is in MB/s, bwlimit in KB/s
- $migrate_speed *= 1024;
-
- $migrate_speed = ($bwlimit < $migrate_speed) ? $bwlimit : $migrate_speed;
-
- # always set migrate speed (overwrite kvm default of 32m) we set a very high
- # default of 8192m which is basically unlimited
- $migrate_speed ||= ($defaults->{migrate_speed} || 8192) * 1024;
-
- # qmp takes migrate_speed in B/s.
- $migrate_speed *= 1024;
- $self->log('info', "migrate_set_speed: $migrate_speed");
- eval {
- PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate_set_speed", value => int($migrate_speed));
- };
- $self->log('info', "migrate_set_speed error: $@") if $@;
-
- my $migrate_downtime = $defaults->{migrate_downtime};
- $migrate_downtime = $conf->{migrate_downtime} if defined($conf->{migrate_downtime});
- if (defined($migrate_downtime)) {
- $self->log('info', "migrate_set_downtime: $migrate_downtime");
- eval {
- PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate_set_downtime", value => int($migrate_downtime*100)/100);
- };
- $self->log('info', "migrate_set_downtime error: $@") if $@;
- }
-
- $self->log('info', "set migration_caps");
- eval {
- PVE::QemuServer::set_migration_caps($vmid);
- };
- warn $@ if $@;
-
- # set cachesize to 10% of the total memory
- my $memory = $conf->{memory} || $defaults->{memory};
- my $cachesize = int($memory * 1048576 / 10);
- $cachesize = round_powerof2($cachesize);
-
- $self->log('info', "set cachesize: $cachesize");
- eval {
- PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate-set-cache-size", value => int($cachesize));
- };
- $self->log('info', "migrate-set-cache-size error: $@") if $@;
-
- if (PVE::QemuServer::vga_conf_has_spice($conf->{vga})) {
- my $rpcenv = PVE::RPCEnvironment::get();
- my $authuser = $rpcenv->get_user();
-
- my (undef, $proxyticket) = PVE::AccessControl::assemble_spice_ticket($authuser, $vmid, $self->{node});
-
- my $filename = "/etc/pve/nodes/$self->{node}/pve-ssl.pem";
- my $subject = PVE::AccessControl::read_x509_subject_spice($filename);
-
- $self->log('info', "spice client_migrate_info");
-
- eval {
- PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "client_migrate_info", protocol => 'spice',
- hostname => $proxyticket, 'tls-port' => $spice_port,
- 'cert-subject' => $subject);
- };
- $self->log('info', "client_migrate_info error: $@") if $@;
-
- }
-
- $self->log('info', "start migrate command to $ruri");
- eval {
- PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate", uri => $ruri);
- };
- my $merr = $@;
- $self->log('info', "migrate uri => $ruri failed: $merr") if $merr;
-
- my $lstat = 0;
- my $usleep = 1000000;
- my $i = 0;
- my $err_count = 0;
- my $lastrem = undef;
- my $downtimecounter = 0;
- while (1) {
- $i++;
- my $avglstat = $lstat/$i if $lstat;
-
- usleep($usleep);
- my $stat;
- eval {
- $stat = PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "query-migrate");
- };
- if (my $err = $@) {
- $err_count++;
- warn "query migrate failed: $err\n";
- $self->log('info', "query migrate failed: $err");
- if ($err_count <= 5) {
- usleep(1000000);
- next;
- }
- die "too many query migrate failures - aborting\n";
- }
-
- if (defined($stat->{status}) && $stat->{status} =~ m/^(setup)$/im) {
- sleep(1);
- next;
- }
-
- if (defined($stat->{status}) && $stat->{status} =~ m/^(active|completed|failed|cancelled)$/im) {
- $merr = undef;
- $err_count = 0;
- if ($stat->{status} eq 'completed') {
- my $delay = time() - $start;
- if ($delay > 0) {
- my $mbps = sprintf "%.2f", $memory / $delay;
- my $downtime = $stat->{downtime} || 0;
- $self->log('info', "migration speed: $mbps MB/s - downtime $downtime ms");
- }
- }
-
- if ($stat->{status} eq 'failed' || $stat->{status} eq 'cancelled') {
- $self->log('info', "migration status error: $stat->{status}");
- die "aborting\n"
- }
-
- if ($stat->{status} ne 'active') {
- $self->log('info', "migration status: $stat->{status}");
- last;
- }
+ livemigrate($self, $vmid, $ruri, $spice_port);
- if ($stat->{ram}->{transferred} ne $lstat) {
- my $trans = $stat->{ram}->{transferred} || 0;
- my $rem = $stat->{ram}->{remaining} || 0;
- my $total = $stat->{ram}->{total} || 0;
- my $xbzrlecachesize = $stat->{"xbzrle-cache"}->{"cache-size"} || 0;
- my $xbzrlebytes = $stat->{"xbzrle-cache"}->{"bytes"} || 0;
- my $xbzrlepages = $stat->{"xbzrle-cache"}->{"pages"} || 0;
- my $xbzrlecachemiss = $stat->{"xbzrle-cache"}->{"cache-miss"} || 0;
- my $xbzrleoverflow = $stat->{"xbzrle-cache"}->{"overflow"} || 0;
- # reduce sleep if remainig memory is lower than the average transfer speed
- $usleep = 100000 if $avglstat && $rem < $avglstat;
-
- $self->log('info', "migration status: $stat->{status} (transferred ${trans}, " .
- "remaining ${rem}), total ${total})");
-
- if (${xbzrlecachesize}) {
- $self->log('info', "migration xbzrle cachesize: ${xbzrlecachesize} transferred ${xbzrlebytes} pages ${xbzrlepages} cachemiss ${xbzrlecachemiss} overflow ${xbzrleoverflow}");
- }
-
- if (($lastrem && $rem > $lastrem ) || ($rem == 0)) {
- $downtimecounter++;
- }
- $lastrem = $rem;
-
- if ($downtimecounter > 5) {
- $downtimecounter = 0;
- $migrate_downtime *= 2;
- $self->log('info', "migrate_set_downtime: $migrate_downtime");
- eval {
- PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate_set_downtime", value => int($migrate_downtime*100)/100);
- };
- $self->log('info', "migrate_set_downtime error: $@") if $@;
- }
-
- }
-
-
- $lstat = $stat->{ram}->{transferred};
-
- } else {
- die $merr if $merr;
- die "unable to parse migration status '$stat->{status}' - aborting\n";
- }
- }
}
sub phase2_cleanup {
@@ -1110,4 +928,190 @@ sub livemigrate_storage {
}
}
+sub livemigrate {
+ my ($self, $vmid, $ruri, $spice_port) = @_;
+
+ $self->log('info', "starting online/live migration on $ruri");
+
+ $self->{livemigration} = 1;
+ my $conf = $self->{vmconf};
+
+ # load_defaults
+ my $defaults = PVE::QemuServer::load_defaults();
+
+ # migrate speed can be set via bwlimit (datacenter.cfg and API) and via the
+ # migrate_speed parameter in qm.conf - take the lower of the two.
+ my $bwlimit = PVE::Storage::get_bandwidth_limit('migrate', undef, $self->{opts}->{bwlimit}) // 0;
+ my $migrate_speed = $conf->{migrate_speed} // $bwlimit;
+ # migrate_speed is in MB/s, bwlimit in KB/s
+ $migrate_speed *= 1024;
+
+ $migrate_speed = ($bwlimit < $migrate_speed) ? $bwlimit : $migrate_speed;
+
+ # always set migrate speed (overwrite kvm default of 32m) we set a very high
+ # default of 8192m which is basically unlimited
+ $migrate_speed ||= ($defaults->{migrate_speed} || 8192) * 1024;
+
+ # qmp takes migrate_speed in B/s.
+ $migrate_speed *= 1024;
+ $self->log('info', "migrate_set_speed: $migrate_speed");
+ eval {
+ PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate_set_speed", value => int($migrate_speed));
+ };
+ $self->log('info', "migrate_set_speed error: $@") if $@;
+
+ my $migrate_downtime = $defaults->{migrate_downtime};
+ $migrate_downtime = $conf->{migrate_downtime} if defined($conf->{migrate_downtime});
+ if (defined($migrate_downtime)) {
+ $self->log('info', "migrate_set_downtime: $migrate_downtime");
+ eval {
+ PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate_set_downtime", value => int($migrate_downtime*100)/100);
+ };
+ $self->log('info', "migrate_set_downtime error: $@") if $@;
+ }
+
+ $self->log('info', "set migration_caps");
+ eval {
+ PVE::QemuServer::set_migration_caps($vmid);
+ };
+ warn $@ if $@;
+
+ # set cachesize to 10% of the total memory
+ my $memory = $conf->{memory} || $defaults->{memory};
+ my $cachesize = int($memory * 1048576 / 10);
+ $cachesize = round_powerof2($cachesize);
+
+ $self->log('info', "set cachesize: $cachesize");
+ eval {
+ PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate-set-cache-size", value => int($cachesize));
+ };
+ $self->log('info', "migrate-set-cache-size error: $@") if $@;
+
+ if (PVE::QemuServer::vga_conf_has_spice($conf->{vga})) {
+ my $rpcenv = PVE::RPCEnvironment::get();
+ my $authuser = $rpcenv->get_user();
+
+ my (undef, $proxyticket) = PVE::AccessControl::assemble_spice_ticket($authuser, $vmid, $self->{node});
+
+ my $filename = "/etc/pve/nodes/$self->{node}/pve-ssl.pem";
+ my $subject = PVE::AccessControl::read_x509_subject_spice($filename);
+
+ $self->log('info', "spice client_migrate_info");
+
+ eval {
+ PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "client_migrate_info", protocol => 'spice',
+ hostname => $proxyticket, 'tls-port' => $spice_port,
+ 'cert-subject' => $subject);
+ };
+ $self->log('info', "client_migrate_info error: $@") if $@;
+
+ }
+
+ $self->log('info', "start migrate command to $ruri");
+ eval {
+ PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate", uri => $ruri);
+ };
+ my $merr = $@;
+ $self->log('info', "migrate uri => $ruri failed: $merr") if $merr;
+
+ my $start = time();
+ my $lstat = 0;
+ my $usleep = 1000000;
+ my $i = 0;
+ my $err_count = 0;
+ my $lastrem = undef;
+ my $downtimecounter = 0;
+ while (1) {
+ $i++;
+ my $avglstat = $lstat/$i if $lstat;
+
+ usleep($usleep);
+ my $stat;
+ eval {
+ $stat = PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "query-migrate");
+ };
+ if (my $err = $@) {
+ $err_count++;
+ warn "query migrate failed: $err\n";
+ $self->log('info', "query migrate failed: $err");
+ if ($err_count <= 5) {
+ usleep(1000000);
+ next;
+ }
+ die "too many query migrate failures - aborting\n";
+ }
+
+ if (defined($stat->{status}) && $stat->{status} =~ m/^(setup)$/im) {
+ sleep(1);
+ next;
+ }
+
+ if (defined($stat->{status}) && $stat->{status} =~ m/^(active|completed|failed|cancelled)$/im) {
+ $merr = undef;
+ $err_count = 0;
+ if ($stat->{status} eq 'completed') {
+ my $delay = time() - $start;
+ if ($delay > 0) {
+ my $mbps = sprintf "%.2f", $memory / $delay;
+ my $downtime = $stat->{downtime} || 0;
+ $self->log('info', "migration speed: $mbps MB/s - downtime $downtime ms");
+ }
+ }
+
+ if ($stat->{status} eq 'failed' || $stat->{status} eq 'cancelled') {
+ $self->log('info', "migration status error: $stat->{status}");
+ die "aborting\n"
+ }
+
+ if ($stat->{status} ne 'active') {
+ $self->log('info', "migration status: $stat->{status}");
+ last;
+ }
+
+ if ($stat->{ram}->{transferred} ne $lstat) {
+ my $trans = $stat->{ram}->{transferred} || 0;
+ my $rem = $stat->{ram}->{remaining} || 0;
+ my $total = $stat->{ram}->{total} || 0;
+ my $xbzrlecachesize = $stat->{"xbzrle-cache"}->{"cache-size"} || 0;
+ my $xbzrlebytes = $stat->{"xbzrle-cache"}->{"bytes"} || 0;
+ my $xbzrlepages = $stat->{"xbzrle-cache"}->{"pages"} || 0;
+ my $xbzrlecachemiss = $stat->{"xbzrle-cache"}->{"cache-miss"} || 0;
+ my $xbzrleoverflow = $stat->{"xbzrle-cache"}->{"overflow"} || 0;
+ # reduce sleep if remainig memory is lower than the average transfer speed
+ $usleep = 100000 if $avglstat && $rem < $avglstat;
+
+ $self->log('info', "migration status: $stat->{status} (transferred ${trans}, " .
+ "remaining ${rem}), total ${total})");
+
+ if (${xbzrlecachesize}) {
+ $self->log('info', "migration xbzrle cachesize: ${xbzrlecachesize} transferred ${xbzrlebytes} pages ${xbzrlepages} cachemiss ${xbzrlecachemiss} overflow ${xbzrleoverflow}");
+ }
+
+ if (($lastrem && $rem > $lastrem ) || ($rem == 0)) {
+ $downtimecounter++;
+ }
+ $lastrem = $rem;
+
+ if ($downtimecounter > 5) {
+ $downtimecounter = 0;
+ $migrate_downtime *= 2;
+ $self->log('info', "migrate_set_downtime: $migrate_downtime");
+ eval {
+ PVE::QemuServer::vm_mon_cmd_nocheck($vmid, "migrate_set_downtime", value => int($migrate_downtime*100)/100);
+ };
+ $self->log('info', "migrate_set_downtime error: $@") if $@;
+ }
+
+ }
+
+
+ $lstat = $stat->{ram}->{transferred};
+
+ } else {
+ die $merr if $merr;
+ die "unable to parse migration status '$stat->{status}' - aborting\n";
+ }
+ }
+}
+
1;
--
2.11.0
More information about the pve-devel
mailing list