[libvirt] [PATCH] qemu: Set correct migrate host in client_migrate_info
https://bugzilla.redhat.com/show_bug.cgi?id=920441 Currently, we are discarding listen attribute from qemu cookie even though we strive to gather it. This result in not so cool bug: if user have different networks, one for management/migration, and one for VNC/SPICE we pass incorrect host to the qemu in client_migrate_info. What we actually pass is remote hostname, while we should be passing remote listen address. It doesn't matter as long as these two are the same, but they don't need necessary to be like that. --- src/conf/domain_conf.c | 4 +++- src/qemu/qemu_migration.c | 2 +- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index e00a532..2afa982 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -16557,8 +16557,10 @@ virDomainGraphicsListenGetAddress(virDomainGraphicsDefPtr def, size_t ii) virDomainGraphicsListenDefPtr listenInfo = virDomainGraphicsGetListen(def, ii, false); + /* even a network can have a listen address */ if (!listenInfo || - (listenInfo->type != VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS)) + !(listenInfo->type == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS || + listenInfo->type == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_NETWORK)) return NULL; return listenInfo->address; } diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 41ad768..089b75a 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -1738,7 +1738,7 @@ qemuDomainMigrateGraphicsRelocate(virQEMUDriverPtr driver, if (ret == 0) { ret = qemuMonitorGraphicsRelocate(priv->mon, cookie->graphics->type, - cookie->remoteHostname, + cookie->graphics->listen, cookie->graphics->port, cookie->graphics->tlsPort, cookie->graphics->tlsSubject); -- 1.8.1.5
On Wed, Apr 10, 2013 at 05:52:26PM +0200, Michal Privoznik wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=920441
Currently, we are discarding listen attribute from qemu cookie even though we strive to gather it. This result in not so cool bug: if user have different networks, one for management/migration, and one for VNC/SPICE we pass incorrect host to the qemu in client_migrate_info. What we actually pass is remote hostname, while we should be passing remote listen address. It doesn't matter as long as these two are the same, but they don't need necessary to be like that. --- src/conf/domain_conf.c | 4 +++- src/qemu/qemu_migration.c | 2 +- 2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index e00a532..2afa982 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -16557,8 +16557,10 @@ virDomainGraphicsListenGetAddress(virDomainGraphicsDefPtr def, size_t ii) virDomainGraphicsListenDefPtr listenInfo = virDomainGraphicsGetListen(def, ii, false);
+ /* even a network can have a listen address */ if (!listenInfo || - (listenInfo->type != VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS)) + !(listenInfo->type == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS || + listenInfo->type == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_NETWORK)) return NULL; return listenInfo->address; } diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 41ad768..089b75a 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -1738,7 +1738,7 @@ qemuDomainMigrateGraphicsRelocate(virQEMUDriverPtr driver, if (ret == 0) { ret = qemuMonitorGraphicsRelocate(priv->mon, cookie->graphics->type, - cookie->remoteHostname, + cookie->graphics->listen,
So how do this work if the user has configured <graphics listen="0.0.0.0"> this is not an address you tell the spice client to reconnect to. This is why it originally used the 'remoteHostname' becuase that is a valid address to connect to. Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|
On 10.04.2013 17:57, Daniel P. Berrange wrote:
On Wed, Apr 10, 2013 at 05:52:26PM +0200, Michal Privoznik wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=920441
Currently, we are discarding listen attribute from qemu cookie even though we strive to gather it. This result in not so cool bug: if user have different networks, one for management/migration, and one for VNC/SPICE we pass incorrect host to the qemu in client_migrate_info. What we actually pass is remote hostname, while we should be passing remote listen address. It doesn't matter as long as these two are the same, but they don't need necessary to be like that. --- src/conf/domain_conf.c | 4 +++- src/qemu/qemu_migration.c | 2 +- 2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index e00a532..2afa982 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -16557,8 +16557,10 @@ virDomainGraphicsListenGetAddress(virDomainGraphicsDefPtr def, size_t ii) virDomainGraphicsListenDefPtr listenInfo = virDomainGraphicsGetListen(def, ii, false);
+ /* even a network can have a listen address */ if (!listenInfo || - (listenInfo->type != VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS)) + !(listenInfo->type == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS || + listenInfo->type == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_NETWORK)) return NULL; return listenInfo->address; } diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 41ad768..089b75a 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -1738,7 +1738,7 @@ qemuDomainMigrateGraphicsRelocate(virQEMUDriverPtr driver, if (ret == 0) { ret = qemuMonitorGraphicsRelocate(priv->mon, cookie->graphics->type, - cookie->remoteHostname, + cookie->graphics->listen,
So how do this work if the user has configured
<graphics listen="0.0.0.0">
this is not an address you tell the spice client to reconnect to. This is why it originally used the 'remoteHostname' becuase that is a valid address to connect to.
Daniel
Aaah, didn't catch that. I suppose in that specific case we should be passing remoteHostname. In such conditions, is 127.0.0.1 specific address as well? I believe so. But I don't see any other specific address. Michal
On Wed, Apr 10, 2013 at 06:01:15PM +0200, Michal Privoznik wrote:
On 10.04.2013 17:57, Daniel P. Berrange wrote:
On Wed, Apr 10, 2013 at 05:52:26PM +0200, Michal Privoznik wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=920441
Currently, we are discarding listen attribute from qemu cookie even though we strive to gather it. This result in not so cool bug: if user have different networks, one for management/migration, and one for VNC/SPICE we pass incorrect host to the qemu in client_migrate_info. What we actually pass is remote hostname, while we should be passing remote listen address. It doesn't matter as long as these two are the same, but they don't need necessary to be like that. --- src/conf/domain_conf.c | 4 +++- src/qemu/qemu_migration.c | 2 +- 2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index e00a532..2afa982 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -16557,8 +16557,10 @@ virDomainGraphicsListenGetAddress(virDomainGraphicsDefPtr def, size_t ii) virDomainGraphicsListenDefPtr listenInfo = virDomainGraphicsGetListen(def, ii, false);
+ /* even a network can have a listen address */ if (!listenInfo || - (listenInfo->type != VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS)) + !(listenInfo->type == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS || + listenInfo->type == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_NETWORK)) return NULL; return listenInfo->address; } diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 41ad768..089b75a 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -1738,7 +1738,7 @@ qemuDomainMigrateGraphicsRelocate(virQEMUDriverPtr driver, if (ret == 0) { ret = qemuMonitorGraphicsRelocate(priv->mon, cookie->graphics->type, - cookie->remoteHostname, + cookie->graphics->listen,
So how do this work if the user has configured
<graphics listen="0.0.0.0">
this is not an address you tell the spice client to reconnect to. This is why it originally used the 'remoteHostname' becuase that is a valid address to connect to.
Daniel
Aaah, didn't catch that. I suppose in that specific case we should be passing remoteHostname. In such conditions, is 127.0.0.1 specific address as well? I believe so. But I don't see any other specific address.
127.0.0.1 is a valid address - it means that any client would have been using SSH tunnelling to connect to it. Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|
participants (2)
-
Daniel P. Berrange -
Michal Privoznik