August 2012 - Devel - libvirt List Archives

[libvirt] [PATCH 1/4] Fix man page file paths to real paths
by Doug Goldstein 15 Aug '12

15 Aug '12

Currently the man page has paths that start with @sysconfdir@, @localstatedir@ and @remote_pid_file@. The sed command attempts to replace these during the build but unfortunately pod2man gets to the files first and escapes the @ character resulting in the sed not working. This removes the @ character and makes the paths correct. --- daemon/Makefile.am | 6 +++--- daemon/libvirtd.pod.in | 18 +++++++++--------- 2 files changed, 12 insertions(+), 12 deletions(-) diff --git a/daemon/Makefile.am b/daemon/Makefile.am index 928aeaf..c15c785 100644 --- a/daemon/Makefile.am +++ b/daemon/Makefile.am @@ -85,9 +85,9 @@ CLEANFILES += test_libvirtd.aug libvirtd.8: $(srcdir)/libvirtd.8.in sed \ - -e 's![@]sysconfdir[@]!$(sysconfdir)!g' \ - -e 's![@]localstatedir[@]!$(localstatedir)!g' \ - -e 's![@]remote_pid_file[@]!$(REMOTE_PID_FILE)!g' \ + -e 's!SYSCONFDIR!$(sysconfdir)!g' \ + -e 's!LOCALSTATEDIR!$(localstatedir)!g' \ + -e 's!REMOTE_PID_FILE!$(REMOTE_PID_FILE)!g' \ < $< > $@-t mv $@-t $@ diff --git a/daemon/libvirtd.pod.in b/daemon/libvirtd.pod.in index 57a3b27..7dc8d83 100644 --- a/daemon/libvirtd.pod.in +++ b/daemon/libvirtd.pod.in @@ -77,14 +77,14 @@ On receipt of B<SIGHUP> libvirtd will reload its configuration. =over -=item F<@sysconfdir@/libvirtd.conf> +=item F<SYSCONFDIR/libvirtd.conf> The default configuration file used by libvirtd, unless overridden on the command line using the B<-f>|B<--config> option. -=item F<@localstatedir@/run/libvirt/libvirt-sock> +=item F<LOCALSTATEDIR/run/libvirt/libvirt-sock> -=item F<@localstatedir@/run/libvirt/libvirt-sock-ro> +=item F<LOCALSTATEDIR/run/libvirt/libvirt-sock-ro> The sockets libvirtd will use when B<run as root>. @@ -92,19 +92,19 @@ The sockets libvirtd will use when B<run as root>. The socket libvirtd will use when run as a B<non-root> user. -=item F<@sysconfdir@/pki/CA/cacert.pem> +=item F<SYSCONFDIR/pki/CA/cacert.pem> The TLS B<Certificate Authority> certificate libvirtd will use. -=item F<@sysconfdir@/pki/libvirt/servercert.pem> +=item F<SYSCONFDIR/pki/libvirt/servercert.pem> The TLS B<Server> certificate libvirtd will use. -=item F<@sysconfdir@/pki/libvirt/private/serverkey.pem> +=item F<SYSCONFDIR/pki/libvirt/private/serverkey.pem> The TLS B<Server> private key libvirtd will use. -=item F<@remote_pid_file@> +=item F<REMOTE_PID_FILE> The PID file to use, unless overridden by the B<-p>|B<--pid-file> option. @@ -121,8 +121,8 @@ To retrieve the version of libvirtd: To start libvirtd, instructing it to daemonize and create a PID file: # libvirtd -d - # ls -la @remote_pid_file@ - -rw-r--r-- 1 root root 6 Jul 9 02:40 @remote_pid_file@ + # ls -la REMOTE_PID_FILE + -rw-r--r-- 1 root root 6 Jul 9 02:40 REMOTE_PID_FILE # =head1 BUGS -- 1.7.8.6

2 4

[libvirt] the arguments of qemudDispatchServerEvent
by mzawdx wang 15 Aug '12

15 Aug '12

Hi all: I am now reading the source codes of libvirt with version 0.9.2 , and be stuck by the function "qemudDispatchServerEvent". as you can see: static void qemudDispatchServerEvent(int watch, int fd, int events, void *opaque) { ... } My question is : except for the argument "void *opaque"， when and how the other three arguments (int watch, int fd, int events) being incoming ? For it is a static function, I look through the source file "libvirt.c" and find no answer . Could someone give me a help ? Thanks a lot. Yours. mzawdx

2 2

[libvirt] Schedule for next release: entering freeze in a week
by Daniel Veillard 15 Aug '12

15 Aug '12

I would rather have the release a bit earlier than expected, I was initially hoping to get a new set of localization for Sep 4 but it is clear at this point that we won't have them in time. Since we have been working on 0.10.0 for 6 weeks now, I think pushing it 2 weeks from now is reasonnable, so I suggest to enter the freeze next week on Wed 22 for a release on the 29 Aug (or a bit earlier if things go well). Considering the amount of changes done for 0.10.0, (we are already at 370 patches in) my guess is that an extra release at the end of September not too long after 0.10.0 but with enough testing feedack will make maintainance easier. So I'm shooting for a 0.10.1 release aproximately 4 weeks after 0.10.0 itself, If such a schedule is raising problem, please tell :-) And if you have patches you want in 0.10.0 which didn't got reviews or isn't pushed, please git rebase them and send them again so they get the attention needed and get pushed within a week ! thanks, Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

2 1

[libvirt] stateless drivers and stored lists of virDomainObjs
by Dmitry Guryanov 15 Aug '12

15 Aug '12

Hello, vmware, openvz and parallels driver, which was written with first two ones as an example, get list of domains in virConnectOpen and then use it for lookups and even for listing sometimes. This is not correct, because you can make several connections to the same driver and changes, made in one connection will not be visible to other. And you can also create or destroy domains using native tools. So I think keeping list of domains in a stateless driver is a bad idea - you should update this list each time when user requests some info. It's better to retrieve needed data only when it requested, like in xenapi driver. I asked this because I want to continue developing parallels driver and it'= s not so hard to change it now, so it should be fixed before adding any new code there. Maybe vmware and openvz drivers should be also fixed. -- Dmitry Guryanov

2 2

[libvirt] [PATCH 1/3] parallels: change URI to parallels:///system
by Dmitry Guryanov 15 Aug '12

15 Aug '12

Let's change URI to parallels:///system. Parallels Server supports creating VMs from non-privileged accounts, but it's not main usage scenario and it may be forbidden in the future. Also containers, which will be supported by the driver, can be managed only by root, so /system path is more suitable for this driver. Signed-off-by: Dmitry Guryanov <dguryanov(a)parallels.com> --- docs/drvparallels.html.in | 10 +++++----- src/parallels/parallels_driver.c | 4 ++-- 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/docs/drvparallels.html.in b/docs/drvparallels.html.in index 40a0fe5..8e1430b 100644 --- a/docs/drvparallels.html.in +++ b/docs/drvparallels.html.in @@ -19,10 +19,10 @@ The libvirt Parallels driver is a single-instance privileged driver, with a driver name of 'parallels'. Some example connection URIs for the libvirt driver are: <pre> -parallels:///default (local access) -parallels+unix:///default (local access) -parallels://example.com/default (remote access, TLS/x509) -parallels+tcp://example.com/default (remote access, SASl/Kerberos) -parallels+ssh://root@example.com/default (remote access, SSH tunnelled) +parallels:///system (local access) +parallels+unix:///system (local access) +parallels://example.com/system (remote access, TLS/x509) +parallels+tcp://example.com/system (remote access, SASl/Kerberos) +parallels+ssh://root@example.com/system (remote access, SSH tunnelled) </pre> </body></html> diff --git a/src/parallels/parallels_driver.c b/src/parallels/parallels_driver.c index 05db54f..53b5a50 100644 --- a/src/parallels/parallels_driver.c +++ b/src/parallels/parallels_driver.c @@ -670,11 +670,11 @@ parallelsOpen(virConnectPtr conn, (conn->uri->path[0] == '/' && conn->uri->path[1] == '\0')) { virReportError(VIR_ERR_INVALID_ARG, "%s", _("parallelsOpen: supply a path or use " - "parallels:///session")); + "parallels:///system")); return VIR_DRV_OPEN_ERROR; } - if (STREQ(conn->uri->path, "/session")) + if (STREQ(conn->uri->path, "/system")) ret = parallelsOpenDefault(conn); else return VIR_DRV_OPEN_DECLINED; -- 1.7.1

2 5

[libvirt] [PATCH] Update documentation for CPU quota/period support for LXC
by Sukadev Bhattiprolu 15 Aug '12

15 Aug '12

From: Sukadev Bhattiprolu <sukadev(a)linux.vnet.ibm.com> Date: Tue, 14 Aug 2012 17:09:11 -0700 Subject: [PATCH] Update documentation for CPU quota/period support for LXC Following commit added suport the CPU quota/period to the LXC driver. Update the documentation to reflect that. commit d9724a81b3c53a40f45bf76067a976cce73ed278 Author: Daniel P. Berrange <berrange(a)redhat.com> Date: Thu Nov 10 12:16:26 2011 +0000 Add support for CPU quota/period to LXC driver Signed-off-by: Sukadev Bhattiprolu <sukadev(a)linux.vnet.ibm.com> --- docs/formatdomain.html.in | 10 ++++++---- 1 files changed, 6 insertions(+), 4 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index ec160bd..fb25cb8 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -426,8 +426,9 @@ interval(unit: microseconds). Within <code>period</code>, each vcpu of the domain will not be allowed to consume more than <code>quota</code> worth of runtime. The value should be in range [1000, 1000000]. A period - with value 0 means no value. (NB: Only qemu driver support) - Since 0.9.4 + with value 0 means no value. + Only QEMU driver support since 0.9.4, LXC since + 0.9.10 </dd> <dt><code>quota</code></dt> <dd> @@ -437,8 +438,9 @@ means that it is not bandwidth controlled. The value should be in range [1000, 18446744073709551] or less than 0. A quota with value 0 means no value. You can use this feature to ensure that all vcpus run at the same - speed. (NB: Only qemu driver support) - Since 0.9.4 + speed. + Only QEMU driver support since 0.9.4, LXC since + 0.9.10 </dd> </dl> -- 1.7.1

2 1

[libvirt] [PATCH 00/48 v3] Atomic APIs to list objects
by Osier Yang 15 Aug '12

15 Aug '12

v2 - v3: * Various document fixes/improvements (Suggested by Laine and Eric) * Destroy virdomainlist.[ch], and folder all list helper funcs into its own *_conf.[ch] * Improve the helpers to get the object list. See https://www.redhat.com/archives/libvir-list/2012-July/msg01267.html * Rebased on top of commit 3df9626 by Peter and the virReportError series by Daniel. * Rebased on top of virsh split series. (I'm luck there is no change on those commands since v2). Except the already supported APIs for domain and domain snapshot, this series add the APIs for the left objects, including storage pool, storage vol, network, interface, node device, nwfilter, and secret. * Storage pool: - Support filtering the returned pool objects by active|inactive, persistent|transient, autostart|no-autostart, and pool types. - New options for virsh, --type to accept multiple pool types. * Storage vol: - Simply returns all the vol objects of a pool. * Network: - Support filtering the results using flags active|inactive, persistent|transient, autostart|no-autostart - New options for virsh. * Interface: - Support filtering the results using flags active|inactive. It's still O(n) underlying, as interface driver doesn't manage the objects itself, but using netcf lib instead. And netcf APIs don't support returning the struct yet. * Node Device: - Support filtering the results using capabilities type of the devices. - Extend --cap to accept multiple capability type. * Network Filter: - Simply returns all the objects. * Secret: - Simply returns all the objects. Osier Yang (48): Avoid strcase check for virsh-*.c Destroy virdomainlist.[ch] list: Define new API virStorageListALlStoragePools list: Add helpers for listing storage pool objects list: Implement the RPC calls for virConnectListAllStoragePools list: Implement listAllStoragePools for storage driver list: Implement listAllStoragePools for test driver list: Add helper to convert strings separated by ', ' to array virsh: Fix the wrong doc for pool-list list: Change MATCH for common use in virsh list: Use virConnectListAllStoragePools in virsh virsh: Use vshPrint instead of printf python: Expose virStorageListAllStoragePools to python binding list: Define new API virStoragePoolListAllVolumes list: Implemente RPC calls for virStoragePoolListAllVolumes list: Implement virStoragePoolListAllVolumes for storage driver list: Implement virStoragePoolListAllVolumes for test driver list: Use virStoragePoolListAllVolumes in virsh list: Expose virStoragePoolListAllVolumes to Python binding list: Define new API virConnectListAllNetworks list: Implement RPC calls for virConnectListAllNetworks list: Add helpers to list network objects list: Implement listAllNetworks for network driver list: Implement listAllNetworks for test driver list: Use virConnectListAllNetworks in virsh list: Expose virConnectListAllNetworks to Python binding list: Define new API virConnectListAllInterfaces list: Implemente RPC calls for virConnectListAllInterfaces list: Implement listAllInterfaces list: Use virConnectListAllInterfaces in virsh list: Expose virConnectListAllInterfaces to Python binding list: Define new API virConnectListAllNodeDevices list: Implemente RPC calls for virConnectListAllNodeDevices list: Add helpers for listing node devices list: Implement listAllNodeDevices list: Expose virConnectListAllNodeDevices to Python binding virsh: Fix a bug of nodedev-list list: Use virConnectListAllNodeDevices in virsh list: Define new API virConnectListAllNWFilters list: Implement RPC calls for virConnectListAllNWFilters list: Implement listAllNWFilters list: Use virConnectListAllNWFilters in virsh list: Expose virConnectListAllNWFilters to Python binding list: Define new API virConnectListAllSecrets list: Implement RPC calls for virConnectListAllSecrets list: Implement listAllSecrets list: Use virConnectListAllSecrets in virsh list: Expose virConnectListAllSecrets to Python binding cfg.mk | 5 +- daemon/remote.c | 382 ++++++++++++++++++++++ include/libvirt/libvirt.h.in | 101 ++++++- python/generator.py | 11 +- python/libvirt-override-api.xml | 44 +++- python/libvirt-override-virConnect.py | 72 +++++ python/libvirt-override-virStoragePool.py | 11 + python/libvirt-override.c | 337 ++++++++++++++++++++ src/Makefile.am | 125 ++++---- src/conf/domain_conf.c | 189 +++++++++++- src/conf/domain_conf.h | 56 ++++ src/conf/network_conf.c | 91 ++++++ src/conf/network_conf.h | 22 ++ src/conf/node_device_conf.c | 103 ++++++ src/conf/node_device_conf.h | 16 + src/conf/storage_conf.c | 116 +++++++ src/conf/storage_conf.h | 35 ++ src/conf/virdomainlist.c | 222 ------------- src/conf/virdomainlist.h | 84 ----- src/driver.h | 35 ++- src/interface/netcf_driver.c | 135 ++++++++ src/libvirt.c | 487 ++++++++++++++++++++++++++++- src/libvirt_private.syms | 4 + src/libvirt_public.syms | 7 + src/libxl/libxl_driver.c | 3 +- src/lxc/lxc_driver.c | 3 +- src/network/bridge_driver.c | 17 + src/node_device/node_device_driver.c | 15 + src/node_device/node_device_driver.h | 3 + src/node_device/node_device_hal.c | 1 + src/node_device/node_device_udev.c | 1 + src/nwfilter/nwfilter_driver.c | 56 ++++ src/openvz/openvz_driver.c | 3 +- src/parallels/parallels_driver.c | 3 +- src/qemu/qemu_driver.c | 3 +- src/remote/remote_driver.c | 449 ++++++++++++++++++++++++++ src/remote/remote_protocol.x | 80 +++++- src/remote_protocol-structs | 85 +++++ src/secret/secret_driver.c | 59 ++++- src/storage/storage_driver.c | 85 +++++ src/test/test_driver.c | 104 ++++++- src/uml/uml_driver.c | 3 +- src/vbox/vbox_tmpl.c | 9 +- src/vmware/vmware_driver.c | 3 +- tools/virsh-domain-monitor.c | 16 +- tools/virsh-domain.c | 19 +- tools/virsh-host.c | 2 +- tools/virsh-interface.c | 259 +++++++++++----- tools/virsh-network.c | 351 ++++++++++++++++----- tools/virsh-nodedev.c | 302 ++++++++++++++++--- tools/virsh-nwfilter.c | 163 ++++++++-- tools/virsh-pool.c | 441 ++++++++++++++++++++------ tools/virsh-secret.c | 182 +++++++++-- tools/virsh-volume.c | 197 +++++++++--- tools/virsh.c | 58 +++- tools/virsh.pod | 51 +++- 56 files changed, 4849 insertions(+), 867 deletions(-) create mode 100644 python/libvirt-override-virStoragePool.py delete mode 100644 src/conf/virdomainlist.c delete mode 100644 src/conf/virdomainlist.h -- 1.7.7.3

4 60

[libvirt] Seamless spice-migration on hosts with multiple interfaces
by Tiziano Müller 15 Aug '12

15 Aug '12

Hi everyone I have a setup in which the VM-nodes have different interfaces for migration (internal network) and public/spice access (public network). Now, when I try to migrate a VM with a connected spice client, the client always gets disconnected instead of seamlessly migrated. The problem is that libvirt passes the spice client the hostname according to the IP of the target interface of the migration instead of the public interface (this has been confirmed using the spicy client which displays an error during migration saying that it can't find the new host "<internal hostname>"). What I tried: * setting spice_listen in /etc/libvirt/qemu.conf * defining a network in libvirt for the public network and using a network in the listen-option for spice in the VMs XML But the problem persists and the client still gets the internal (migration) address instead of the public where spice is listening on. This is with libvirt 0.9.13. Has anyone seen this behaviour before? I tried to track down this problem and I'm not sure whether I find one bug already, but the following code didn't look right. So I tried to fix it (but the problem was still present): --- src/qemu/qemu_migration.c.orig 2012-07-20 16:46:59.669855341 +0200 +++ src/qemu/qemu_migration.c 2012-07-20 16:47:58.637546032 +0200 @@ -1036,7 +1036,7 @@ if (ret == 0) { ret = qemuMonitorGraphicsRelocate(priv->mon, cookie->graphics->type, - cookie->remoteHostname, + cookie->graphics->listen, cookie->graphics->port, cookie->graphics->tlsPort, cookie->graphics->tlsSubject); Thanks in advance, Tiziano

1 0

[libvirt] [PATCHv2 0/5] network: report #connections in network xml
by Laine Stump 15 Aug '12

15 Aug '12

This repost of an unreviewed patchset from last week is purely to rebase - the "openvswitch in <network> 9 patch series is a prerequisite of this, and this is a prerequisite of the vlan patch series resend coming up. These are all fairly short and simple, so shouldn't require too much hair pulling to review :-) ===== The forward devices in direct mode (aka macvtap) networks already have a counter to keep track of how many guest interfaces are using each physical device. This series renames that counter to "connections", then adds another connects counter to the toplevel of each network, and outputs both of these counters (when non-0) with the live XML for the network (i.e. it isn't output if you ask for the --inactive xml). Someone working on a management application had requested this info, and it already existed, so I thought I may as well expose it.

3 22

[libvirt] [PATCH] conf: log error when encountering multiple filterrefs in interface
by Laine Stump 15 Aug '12

15 Aug '12

Each interface has a single pointer to a filterref object. That filterref can itself point to multiple other filterrefs, but at the toplevel there is only one. The parser had previously just silently overwritten earlier filterrefs when a new one was encountered, so the interface was left with whichever was the last filterref in the xml, ignoring all the others. This patch logs an error when it sees more than one filterref. --- src/conf/domain_conf.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index ae95d81..58c424f 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -4627,6 +4627,12 @@ virDomainNetDefParseXML(virCapsPtr caps, ioeventfd = virXMLPropString(cur, "ioeventfd"); event_idx = virXMLPropString(cur, "event_idx"); } else if (xmlStrEqual (cur->name, BAD_CAST "filterref")) { + if (filter) { + virReportError(VIR_ERR_XML_ERROR, "%s", + _("Invalid specification of multiple <filterref>s " + "in a single <interface>")); + goto error; + } filter = virXMLPropString(cur, "filter"); virNWFilterHashTableFree(filterparams); filterparams = virNWFilterParseParamAttributes(cur); -- 1.7.11.2

2 2