- Devel - libvirt List Archives

[Libvir] Libvirt dependencies
by Karthik V 13 Nov '06

13 Nov '06

I'm trying to write a Java wrapper around some libvirt functions using Java Native Interface (JNI). To try this out, I tried writing JNI code that invokes the getDomainInfo() function in the example program info1.c (given with the distribution). While trying to execute the Java wrapper that invokes this C function, I encounter this error: Inconsistency detected by ld.so: dl-deps.c: 559: _dl_map_object_deps: Assertion`map->l_searchlist.r_list[0] == map' failed! I suspect this is due to some missing library that libvirt depends on. As part of the JNI code compilation, I'm supposed to create a shared library that includes all the dependent libraries. Currently, I'm using this command line to do that: ld -shared -f /usr/local/lib/libvirt.so -lxenstore -lpthread -lz LibvirtWrapper.o -o libjava-libvirt.so Here, I assume that libvirt.so and libxenstore are the only dependencies. Can someone tell me if I'm missing some other library in the above command line?

3 2

Re: [Libvir] First virConnectOpen() failed
by Glen Deem 10 Nov '06

10 Nov '06

Thanks, Daniel, I can see virsh nodeinfo (custom) passes now, but I still get the same error: when running under my user account: PASS: virshtest confdata/fc4.conf OK PASS: test_conf.sh First virConnectOpen() failed FAIL: reconnect when running as root: PASS: virshtest confdata/fc4.conf OK PASS: test_conf.sh First lookup for domain 0 failed FAIL: reconnect I suspect there is something wrong with my Xen installation. What to you think? Thanks, Glen Daniel Veillard wrote: > On Thu, Nov 09, 2006 at 03:41:55PM -0500, Glen Deem wrote: >> Hi all, >> >> I've compiled libvirt from the head, running tests failed with: >> >> first virConnectOpen() failed >> >> I've traced it to a failure of xs_daemon_open_readonly() that returns 0. > > it was a test problem when running as non-root, fixed in CVS yesterday :-) > > Daniel >

2 1

[Libvir] Failed to find domain 0 error
by Raja 10 Nov '06

10 Nov '06

Dear friends, i installed libvirt 0.1.8 on my xen 3.0.3 witout any trouble. I also installed the python bindings. I am able to run the python script (basic.py) to get information about domain0 but running info1 in examples in docs does not work. When i run it , i get " Failure to find domain0 "as output. I am running info1 in domain0. Why is this?. is this a bug?. Please help me. Thanks, Raja ____________________________________________________________________________________ Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com

2 1

[Libvir] virt-manager 0.2.6 release
by Daniel P. Berrange 10 Nov '06

10 Nov '06

I'm pleased to announce the 0.2.6 release of 'Virtual Machine Manager' aka 'virt-manager' is available for download: http://virt-manager.et.redhat.com/download.html The release focus has been on major bug fixing. This is also the first release in which (partial) translations are available for the UI in approximately 20 languages - thanks to the Fedora i18n team for excellant progress on this. It is now possible to control the virt-manager UI with command line arguments as well as the DBus API & DBus activation is no longer used by default which fixes interaction with GNOME keyring & AT-SPI accesibility. Numerous UI issues were fixed / clarified, particularly in the graphical console and new VM creation wizard. There are updated screenshots of the UI on the project website: http://virt-manager.et.redhat.com/screenshots.html As ever, report bugs against the 'virt-manager' component in the Red Hat Bugzilla server. Work is also underway to address the frequently requested feature of management for inactive domains. Since this is not supported by XenD on any current release we're aiming to add support for loading config files straight out of /etc/xen into libvirt & virt-manager. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

1 0

[Libvir] First virConnectOpen() failed
by Glen Deem 09 Nov '06

09 Nov '06

Hi all, I've compiled libvirt from the head, running tests failed with: first virConnectOpen() failed I've traced it to a failure of xs_daemon_open_readonly() that returns 0. What would be a problem, please? xend is running, I see dom0 and can create domUs etc... I use Debian unstable Xen packages. Thanks in advance, Glenn P.S. I'm not a Xen expert, kind of intermediate/advanced user.

2 1

[Libvir] compatibility issues
by Raja 07 Nov '06

07 Nov '06

Dear friends, i tried using libvirt on xen3.0 on a debian machine. When i invoked make, i got a lot of error messages. i went through some posts and i found out that libvirt was not compatible with certain versions of xen. I would like to know as to which versions of xen support libvirt so taht i can start working on them.. Thanks, Raja --------------------------------- Sponsored Link Talk more and pay less. Vonage can save you up to $300 a year on your phone bill. Sign up now.

2 1

[Libvir] Configure error - Could not find libxml2 anywhere
by Karthik V 07 Nov '06

07 Nov '06

Hi All, I'm new to libvirt, and am having trouble running configure script. This is the error - checking for library containing xs_read... -lxenstore checking libxml2 no ... checking for libxml2 libraries >= 2.5.0... checking libxml2 xml2-config ... configure: error: Could not find libxml2 anywhere (see config.log for details). I've verified that I have libxml2 installed - debian:~/libvirt-0.1.8# dpkg -l | grep libxml ii libxml2 2.6.16-7 GNOME XML library Can someone help me fix this problem? Thanks Karthik

2 6

[Libvir] virt-manager command line options
by Karel Zak 26 Oct '06

26 Oct '06

We need a way how call virt-manager methods without D-BUS. This patch adds new --show-* command line options. The --show-* options need to be used together with the --connect option. Daniel, I also added an OptionValueError exception. It's not perfect solution, but I'm not sure what's your favourite way for error messages in virt-manager when GTK is not initialized yet. # virt-manager --help usage: virt-manager.py [options] options: -h, --help show this help message and exit -c URI, --connect=URI Connect to hypervisor at URI --no-dbus Disable DBus service for controlling UI --show-domain-creator Create a new virtual machine --show-domain-editor=UUID Edit a domain configuration --show-domain-performance=UUID Show a domain performance --show-domain-console=UUID Show a domain console --show-host-summary Show a host summary Karel # HG changeset patch # User Karel Zak <kzak(a)redhat.com> # Node ID 3aab017160416137fe3033e13a206303287e5dd9 # Parent 127d61c9062e0ed5662f412f5edbb6c7dbd5582f add command line option diff -r 127d61c9062e -r 3aab01716041 src/virt-manager.py.in --- a/src/virt-manager.py.in Thu Oct 19 13:48:03 2006 -0400 +++ b/src/virt-manager.py.in Wed Oct 25 23:24:24 2006 +0200 @@ -70,7 +70,7 @@ dbus.glib.threads_init() dbus.glib.threads_init() import dbus.service -from optparse import OptionParser +from optparse import OptionParser, OptionValueError appname = "::PACKAGE::" appversion = "::VERSION::" @@ -102,21 +102,74 @@ from virtManager.remote import vmmRemote gtk.window_set_default_icon_from_file(icon_dir + "/" + appname + "-icon.svg") +# maps --show-* to engine methods +def show_engine(engine, show, uri, uuid): + if show=='creator': + engine.show_create(uri) + elif show=='editor': + engine.show_details_config(uri, uuid) + elif show=='performance': + engine.show_details_performance(uri, uuid) + elif show=='console': + engine.show_console(uri, uuid) + elif show=='summary' or uri: + engine.show_manager(uri) + else: + engine.show_connect() + +# maps --show-* to remote manager methods +def show_remote(managerObj, show, uri, uuid): + if show=='creator': + managerObj.show_domain_creator(uri) + elif show=='editor': + managerObj.show_domain_editor(uri, uuid) + elif show=='performance': + managerObj.show_domain_performance(uri, uuid) + elif show=='console': + managerObj.show_domain_console(uri, uuid) + elif show=='summary' or uri: + managerObj.show_host_summary(uri) + else: + managerObj.show_connect() + +# Generic OptionParser callback for all --show-* options +# This routine stores UUID to options.uuid for all --show-* options +# where is metavar="UUID" and also sets options.show +def opt_show_cb(option, opt_str, value, parser): + if option.metavar=="UUID": + setattr(parser.values, "uuid", value) + s = str(option) + show = s[s.rindex('-')+1:] + setattr(parser.values, "show", show) + # Run me! def main(): optParser = OptionParser() - optParser.add_option("-c", "--connect", dest="uri", help="Connect to hypervisor at URI", metavar="URI") - optParser.add_option("--no-dbus", action="store_true", dest="nodbus", help="Disable DBus service for controlling UI") + optParser.set_defaults(uuid=None) + optParser.add_option("-c", "--connect", dest="uri", + help="Connect to hypervisor at URI", metavar="URI") + optParser.add_option("--no-dbus", action="store_true", dest="nodbus", + help="Disable DBus service for controlling UI") + optParser.add_option("--show-domain-creator", action="callback", + callback=opt_show_cb, dest="show", help="Create a new virtual machine") + optParser.add_option("--show-domain-editor", type="string", metavar="UUID", + action="callback", callback=opt_show_cb, help="Edit a domain configuration") + optParser.add_option("--show-domain-performance", type="string", metavar="UUID", + action="callback", callback=opt_show_cb, help="Show a domain performance") + optParser.add_option("--show-domain-console", type="string", metavar="UUID", + action="callback", callback=opt_show_cb, help="Show a domain console") + optParser.add_option("--show-host-summary", action="callback", + callback=opt_show_cb, help="Show a host summary") (options, args) = optParser.parse_args() + + if options.show and options.uri==None: + raise OptionValueError("can't use --show-* options without --connect") config = vmmConfig(appname, appversion, gconf_dir, glade_dir, icon_dir) engine = vmmEngine(config) if options.nodbus: - if options.uri != None: - engine.show_manager(options.uri) - else: - engine.show_connect() + show_engine(engine, options.show, options.uri, options.uuid) else: bus = None try: @@ -135,19 +188,14 @@ def main(): managerProxy = bus.get_object("com.redhat.virt.manager", "/com/redhat/virt/manager") managerObj = dbus.Interface(managerProxy, "com.redhat.virt.manager") - if options.uri != None: - managerObj.show_host_summary(options.uri) - else: - managerObj.show_connect() + show_remote(managerObj, options.show, options.uri, options.uuid) + # yes, we exit completely now - remote service is in charge return except: logging.warning("Could not connection to session bus, disabling DBus service " + \ str(sys.exc_info()[0]) + " " + str(sys.exc_info()[1])) - if options.uri != None: - engine.show_manager(options.uri) - else: - engine.show_connect() + show_engine(engine, options.show, options.uri, options.uuid) gtk.main() if __name__ == "__main__":

2 1

[Libvir] Include actual error info in python exceptions
by Daniel P. Berrange 25 Oct '06

25 Oct '06

The current python bindings throw a generic libvirtError exception object whose message merely consists of the name of the method which failed. This isn't really all that useful to the caller, since they know exactly which method failed - its the one they just called! The attached patch adds a binding for virConGetLastError & virGetLastError & pulls this info into the libvirtError object being thrown. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 1

[Libvir] Ensure errors are reported against virConnect object
by Daniel P. Berrange 25 Oct '06

25 Oct '06

In xend_internal.c there are a large number of places where errors are being reported without supplying an associated virConnectPtr object. So for example when calling virDomainCreateLinux, some errors get associated with the connection object, while others get attached to the global error object. The attached patch fixes up all places in xend_internal.c which have an virConnectPtr object available to always pass this object into the the error reporting funcs. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 1

[Libvir] Make python binding release GIL when using threads
by Daniel P. Berrange 24 Oct '06

24 Oct '06

In the virt-manager app I'm using the python bindings to libvirt for all operations. For long running operations such as creating new domains, or saving/restoring domains I fork a background thread to do the call, so that the main UI / status refresh can carry on working. Except this wasn't working. The main thread was being blocked for the entire duration of the background thread. After a lot of nasty debug sessions I discovered that the main thread was being blocked in the GTK event loop attempting to aquire the Python GIL (Global Interpreter Lock). The only way this could be occurring is if the background thread doing the libvirt call held the GIL. And indeed after looking at the source for the libvirt python bindings it appears we never release the GIL when calling into the libvirt C apis, so for the entire duration of virDomainCreateLinux the python interpreter is completely stopped. Clearly this sucks because virDomainCreateLinux can take a very long time. I read up a little on python threading & looked at how PyGTK deals with this & have come up with a patch to make libvirt release the GIL when entering a C call & re-aquire it on completion. Indeed, since PyGTK & GTK are both LGPL, I just copied their C macros straight over. The patch appears to work - the python interpreter is no longer deadlocked when calling virDomainCreateLinux (or any other libvirt calls). Unfortuntely the virtmanager app still deadlocks, because it turns out XenD itself serializes all SEXPR HTTP requests, but once this is fixed the whole thing is fully parallelized & works without blocking the app. Attached the patch - I've not done much python threading before so let me know if i missed anything / messed stuff up. NB, I also hacked up the error handling callback so it re-acquires the lock before calling back into the python. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 2

[Libvir] Release of libvirt-0.1.8
by Daniel Veillard 16 Oct '06

16 Oct '06

There have been a number of bug raised and fixed since the 0.1.7 release: * Bug for system with page size != 4k * vcpu number initialization (Philippe Berthault) * don't label crashed domains as shut off (Peter Vetere) * fix virsh man page (Noriko Mizumoto) * blktapdd support for alternate drivers like blktap (Daniel Berrange) * memory leak fixes (xend interface and XML parsing) (Daniel Berrange) * compile fix * mlock/munlock size fixes (Daniel Berrange) * improve error reporting The new package is available as usual at ftp://libvirt.org/libvirt/ thanks to everybody who helped by reporting or fixing bugs ! Daniel -- Red Hat Virtualization group http://redhat.com/virtualization/ Daniel Veillard | virtualization library http://libvirt.org/ veillard(a)redhat.com | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/

1 0

[Libvir] Potential problem on NUMA machines
by Philippe Berthault 13 Oct '06

13 Oct '06

In the virNodeInfo structure, it's assumed that all nodes of multiprocessor systems have the same number of sockets. But on NUMA machines, it's not mandatory that all SMP nodes have the same number of sockets. So there is a potential problem when computing the number of processors by using the informations returned in the virNodeInfo structure. This problem exists also in Xen (in the "xm info" command) but both Xen and libvirt are wrong.

2 1

[Libvir] virDomainGetInfo returns a false result on IA64 systems
by Philippe Berthault 12 Oct '06

12 Oct '06

The maxMem and memory fields of the virDomainInfo structure are false on IA64 systems. The problem is in xen_internal.c at lines 1530,1531: info->memory = XEN_GETDOMAININFO_TOT_PAGES(dominfo) * 4; info->maxMem = XEN_GETDOMAININFO_MAX_PAGES(dominfo) * 4; It's assumed that page size is 4 KB (* 4) but this isn't correct on a IA64 system where the page size is variable. I haven't checked but I suppose this bug exists also with the virDomainGetMaxMemory and virDomainSetMaxMemory functions. Regards.

3 4

[Libvir] Bug correction in virDomainGetVcpus() function
by Philippe Berthault 12 Oct '06

12 Oct '06

In the virXen_getvcpusinfo() function of xen_internal.c source file, the vcpu number stored in the virVcpuInfo structure is the xen domain id instead of the vcpu id. Here the patch of the correction. --- libvirt-0.1.7/src/xen_internal.c 2006-09-29 10:18:17.000000000 +0200 +++ libvirt-0.1.8/src/xen_internal.c 2006-10-12 11:29:40.000000000 +0200 @@ -1012,7 +1012,7 @@ ret = xenHypervisorDoV2Dom(handle, &op); if (ret < 0) return(-1); - ipt->number = id; + ipt->number = vcpu; if (op.u.getvcpuinfo.online) { if (op.u.getvcpuinfo.running) ipt->state = VIR_VCPU_RUNNING; if (op.u.getvcpuinfo.blocked) ipt->state = VIR_VCPU_BLOCKED; @@ -1054,7 +1054,7 @@ ret = xenHypervisorDoV1Op(handle, &op); if (ret < 0) return(-1); - ipt->number = id; + ipt->number = vcpu; if (op.u.getvcpuinfo.online) { if (op.u.getvcpuinfo.running) ipt->state = VIR_VCPU_RUNNING; if (op.u.getvcpuinfo.blocked) ipt->state = VIR_VCPU_BLOCKED; @@ -1078,7 +1078,7 @@ ret = xenHypervisorDoV0Op(handle, &op); if (ret < 0) return(-1); - ipt->number = id; + ipt->number = vcpu; if (op.u.getvcpuinfo.online) { if (op.u.getvcpuinfo.running) ipt->state = VIR_VCPU_RUNNING; if (op.u.getvcpuinfo.blocked) ipt->state = VIR_VCPU_BLOCKED;

2 1

[Libvir] [PATCH] libvirt incorrectly reports crashed domains as shutoff
by Peter Vetere 11 Oct '06

11 Oct '06

Hi, I've noticed that libvirt seems to report crashed domains as shutoff. Here's an example: [root@XXXXX ~]# virsh list Id Name State ---------------------------------- 0 Domain-0 running 27 test03 blocked 34 test01 shut off 36 onetwentyeight blocked [root@XXXXX ~]# xm list Name ID Mem(MiB) VCPUs State Time(s) Domain-0 0 356 1 r----- 3655.4 onetwentyeight 36 128 1 -b---- 16.7 test01 34 256 1 ----c- 7.4 test03 27 256 1 -b---- 231.7 The reason is because xen_internal.c does not interpret the additional shutdown flags when it checks the domain's state. Attached is a patch which correctly checks these flags and reports the crashed state. I wasn't sure of the best way to react to the other shutdown flags, so I just defaulted to "SHUTOFF" for everything else. Pete

3 2

[Libvir] Supporting new Xen 3.0.3 blktap backend for file devices
by Daniel P. Berrange 06 Oct '06

06 Oct '06

When defining a guest's block device in the libvirt XML we currently allow the dev to be backed by either a file or block device in the host system. This is done by specifying the 'type' attribute when defing the block device XML. eg <disk type='file'> Or <disk type='block'> This is all fine & dandy for Xen 3.0.2, but in 3.0.3 we now have a choice of drivers for the 'file' backed block devices. We can either use the traditional loopback mounted system, or the new blktap system. To complicate things further still, blktap itself has man different drivers - one uses a raw file with Async-IO, another uses the QCow sparse format with AIO, and all sort of other possibilities are available... This means the easy option of just making all file devices use blktap in 3.0.3 is not practical. This in turns means we need to come up with a way to express the new driver methods in libvirt XML. There are a few options I can think of : - Allow more values in the 'type' attribute, eg file, block, blktap:aio, blktap:qcow, etc. This feels wrong because the blktap:* entries are really still 'file' types. - Introduce a new attribute 'driver' where you can specify 'loop', 'blktap:aio', 'blktap:qcow', etc - Introduce a new element 'driver' where you can specify the same <disk type="file"> <driver type='blktap:aio'/> </disk> - Same as above, but normalize the driver type further <disk type="file"> <driver type='blktap' backend='aio' /> </disk> My preference is probably option 3 or 4. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

3 7

Re: [Libvir] Xen Routed
by Daniel P. Berrange 04 Oct '06

04 Oct '06

On Wed, Oct 04, 2006 at 08:09:27AM -0600, Nick Devito wrote: > On Wed, 2006-10-04 at 14:06 +0100, Daniel P. Berrange wrote: > > On Tue, Oct 03, 2006 at 07:15:23PM -0700, Sargun Dhillon wrote: > > > Does the XML file format support routed? > > > If so, what would the <interface> clause look like? > > > > I'm not familiar with the way routed networking is setup with Xen. If you > > can provide the example VIF device config params from /etc/xen/<some guest> > > which are needed for routing I'll check if they map through to libvirt > > XML format in neccessary manner. > > > Xen routed requires the ip= configuration variable, and, I think thats > it. I don't see any reason why it shouldn't work correctly. Ok, looking at the XML parsing routines we don't currently support the 'ip' property. Fortunately, it looks as if one of Michel Ponceau's patches from a thread last week ought to solve this problem, so we ought to be able to get this working for next release. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

1 0

[Libvir] Xen Routed
by Sargun Dhillon 04 Oct '06

04 Oct '06

Does the XML file format support routed? If so, what would the <interface> clause look like? -- Sargun Dhillon President Atarack Communications, Inc. (925)-202-9485

2 1

[Libvir] Problem with hypercall
by Jim Fehlig 02 Oct '06

02 Oct '06

Stumbled across a problem trying to list domains with libvirt-0.1.5. Using virsh I get errors such as xen81:/tests/jim # virsh list Id Name State ---------------------------------- 0 Domain-0 running libvir: Xen Daemon error : GET operation failed: No such domain 16 xm shows xen81:/tests/jim # xm list Name ID Mem(MiB) VCPUs State Time(s) Domain-0 0 2048 4 r----- 815825.8 vm1 1 512 2 -b---- 0.3 I found that the buffer provided for XEN_V1_OP_GETDOMAININFOLIST hypercall differs slightly from the buffer in xen/dom0_ops.h. Attached is a patch against current cvs that works for me, but I'm not familiar with this part of the code so not sure if this is the proper fix. I'm using xen 3.0.2 that shipped with SLES10. Regards, Jim

3 8

[Libvir] Proposal for virtual devices described in XML
by michel.ponceau＠bull.net 02 Oct '06

02 Oct '06

Proposal modified after comments from Daniel P. Berrange: /* Create a virtual device attachment to backend */ int virDomainAttachDevice(virDomainPtr domain, char *xml); /* Destroy a virtual device attachment to backend */ int virDomainDetachDevice(virDomainPtr domain, char *xml); /* @domain: pointer to domain object * @xml: pointer to XML description of one device * Returns 0 in case of success, -1 in case of failure. */ The XML would be in same form as the part of domain description for a single device, either <disk.../disk> or <interface.../interface>. In Detach function, the device is identified by its target name for "disk", and its MAC address for "interface". When the MAC address has not been assigned by Libvirt user at device creation, it must be retrieved from Xen using virDomainGetXMLDesc function.

1 0

[Libvir] Release of libvirt-0.1.7
by Daniel Veillard 29 Sep '06

29 Sep '06

Is available on ftp://libvirt.org/libvirt/ It's 0.1.6 with the fixes for the 2 bugs that Daniel Berrange fixed yesterday: - failure of domain info when using the hypervisor on 32 bits with 3.0.3 - the memory corrption which can happen when reading vCPU info from xend Daniel -- Red Hat Virtualization group http://redhat.com/virtualization/ Daniel Veillard | virtualization library http://libvirt.org/ veillard(a)redhat.com | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/

1 0

Re: [Libvir] Proposal for virtual devices described in XML
by michel.ponceau＠bull.net 28 Sep '06

28 Sep '06

When specified at device creation, vifname is put in XenStore directory /local/domain/0/backend/vif/<domID>/<vif_number>. All vifnames used in a domain are returned by Get "/xend/domain/<domID>?detail=1" and converted into XML "target" by virDomainGetXMLDesc. For convenience, we could force the vifname, either at domain creation or in Attach, and ignore any "target" in "interface", except for Detach. The vifname value could be the sequence number of network device creation in domain (for Xen, it would be equal to vif number). Veuillez répondre à "Daniel P. Berrange" <berrange(a)redhat.com> Pour : michel.ponceau(a)bull.net cc : libvir-list(a)redhat.com Objet : Re: [Libvir] Proposal for virtual devices described in XML On Wed, Sep 27, 2006 at 05:21:33PM +0200, michel.ponceau(a)bull.net wrote: > After comments from Daniel P. Berrange and from Bull people, my proposal > is reduced to following 2 functions: > > /* Create a virtual device attachment to backend */ > int virDomainAttachDevice(virDomainPtr domain, char *xml); This is definitely a good way forward - it lets us re-use the existing code for parsing device descriptions very easily. > /* Destroy a virtual device attachment to backend */ > int virDomainDetachDevice(virDomainPtr domain, char *xml); > /* @domain: pointer to domain object > * @xml: pointer to XML description of one device > * Returns 0 in case of success, -1 in case of failure. */ I was wondering whether we need to have the full XML block passed in for detaching devices. Perhaps it is sufficient to just pass i nthe target name of the device. Then again, some implementations may prefer to have the full XML blob available to allow greater sanity checking when detaching, so its probably worth while. > The XML would be in same form as the part of domain description for a > single device, either <disk.../disk> or <interface.../interface>. Very nice. > The device is identified by its target name. This attribute is optional > for "interface" in domain description, but would become mandatory in > Attach/Detach description. It is not yet processed in Libvirt-0.1.6 source > xml.c (function virDomainParseXMLIfDesc). I have successfully tried the attached patch, which tranlates "target" > into Xen "vifname". (It processes "ip" attribute in same way, but my try > failed, maybe because address value was invalid for Xen). I'm not sure I like the idea of having to explicitly set a vif name for every guest NIC we define - principally because there's no straightforward way for a user of this API to determine what VIF names are already in use, vs free. It also seems XenD does not tell us what the vifname for running domains are - although I think the info is in xenstore somewhere. I can see why we need the vifname to detach the device though. If we can find a way to fill in the vifname when dumping the XML, then this would let a caller uniquely identify an existing device, without us requireing that vifnames are used when creating a domain/attaching a new device (& thus let Xen still auto-allocate them) Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 1

[Libvir] Proposal for virtual devices described in XML
by michel.ponceau＠bull.net 28 Sep '06

28 Sep '06

My try of "ip" attribute failed because of a bug in patch. With this corrected file, it works as well as "target". Pour : libvir-list(a)redhat.com cc : Objet : Proposal for virtual devices described in XML After comments from Daniel P. Berrange and from Bull people, my proposal is reduced to following 2 functions: /* Create a virtual device attachment to backend */ int virDomainAttachDevice(virDomainPtr domain, char *xml); /* Destroy a virtual device attachment to backend */ int virDomainDetachDevice(virDomainPtr domain, char *xml); /* @domain: pointer to domain object * @xml: pointer to XML description of one device * Returns 0 in case of success, -1 in case of failure. */ The XML would be in same form as the part of domain description for a single device, either <disk.../disk> or <interface.../interface>. The device is identified by its target name. This attribute is optional for "interface" in domain description, but would become mandatory in Attach/Detach description. It is not yet processed in Libvirt-0.1.6 source xml.c (function virDomainParseXMLIfDesc). I have successfully tried the attached patch, which tranlates "target" into Xen "vifname". (It processes "ip" attribute in same way, but my try failed, maybe because address value was invalid for Xen).

1 0

[Libvir] virDomainDestroy not synchronous?
by Lon Hohberger 28 Sep '06

28 Sep '06

i686 libvirt-0.1.5-3 xen-3.0.2-33 Code: domain_desc = virDomainGetXMLDesc(vdp, 0); response = virDomainDestroy(vdp); if (response == 0 && domain_desc) { /* sleep(3); */ virDomainCreateLinux(vp, domain_desc, 0); free(domain_desc); } If I add the 'sleep()', the domain dies and virDomainCreateLinux() fails with the following error: libvir: Xen Daemon error : POST operation failed: No such domain futon1 Failed to get devices for domain futon1 That really doesn't bother me, and it's likely my fault anyway (it's not a fully configured domain; it's a dom-U sitting in the anaconda install screen). Of course, it's moot if we get a guaranteed synchronous reboot flag for virDomainReboot()... hint hint ;) What bothers me is that if I *don't* put the sleep in, the domain never actually gets destroyed. In this particular case, the domain is still sitting at the anaconda install screen; all I have to do is reconnect using xm console. If I only do a virDomainDestroy(), the domain gets destroyed. So, it looks like like the virDomainCreateLinux() cancels the destroy request before it can be completed. This makes it look like that the destroy request is not synchronous, and/or the result of the operation is not guaranteed (only that the request was made). Is it supposed to work this way, or is this a bug/problem? -- Lon

4 7

[Libvir] Proposal for virtual devices described in XML
by michel.ponceau＠bull.net 27 Sep '06

27 Sep '06

After comments from Daniel P. Berrange and from Bull people, my proposal is reduced to following 2 functions: /* Create a virtual device attachment to backend */ int virDomainAttachDevice(virDomainPtr domain, char *xml); /* Destroy a virtual device attachment to backend */ int virDomainDetachDevice(virDomainPtr domain, char *xml); /* @domain: pointer to domain object * @xml: pointer to XML description of one device * Returns 0 in case of success, -1 in case of failure. */ The XML would be in same form as the part of domain description for a single device, either <disk.../disk> or <interface.../interface>. The device is identified by its target name. This attribute is optional for "interface" in domain description, but would become mandatory in Attach/Detach description. It is not yet processed in Libvirt-0.1.6 source xml.c (function virDomainParseXMLIfDesc). I have successfully tried the attached patch, which tranlates "target" into Xen "vifname". (It processes "ip" attribute in same way, but my try failed, maybe because address value was invalid for Xen).

2 1

[Libvir] Flag to virDomainReboot to immediately reboot a domain?
by Lon Hohberger 22 Sep '06

22 Sep '06

Hi, For fencing Xen domains, I would like the ability to guarantee an immediate reboot of a domain without notifying the OS inside. Basically, I would like something equivalent to pressing the reset button on a physical machine. There's a currently unused "flags" option to virDomainReboot which would allow such an addition without breaking the ABI. Doing getXMLDesc/Destroy/Create introduces a race condition in clusters: What if outside management software notices notices the VM is missing and decides to start it up on *another* host node before we bring up the VM again using the Create call? An immediate reboot should eliminate the race (even though it's a very small window), and probably will not add any complexity. For now, 'reboot' is done as above, and 'off' is a simple Destroy call. 'On' is not possible to implement for various reasons, the most obvious of which is that the caller does not know any of the configuration information about particular domUs (except for their names). -- Lon

1 0

[Libvir] Release of libvirt-0.1.6
by Daniel Veillard 22 Sep '06

22 Sep '06

Available at ftp://libvirt.org/libvirt/ Mostly a few bug fixes, preparing the upcoming xen-3.0.3, and localization support were added in the last 2 weeks: * Support for localization of strings using gettext (Daniel Berrange) * Support for new Xen-3.0.3 cdrom and disk configuration (Daniel Berrange) * Support for setting VNC port when creating domains with new xend config files (Daniel Berrange) * Fix bug when running against xen-3.0.2 hypercalls (Jim Fehlig) * Fix reconnection problem when talking directly to http xend thanks to everybody who helped testing, improving and fixing it ! Daniel -- Red Hat Virtualization group http://redhat.com/virtualization/ Daniel Veillard | virtualization library http://libvirt.org/ veillard(a)redhat.com | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/

1 0

[Libvir] gettext i18n support for libvirt
by Daniel P. Berrange 21 Sep '06

21 Sep '06

Attached is an updated version of my previous patch to add gettext support to libvirt. It should be pretty self-explanatory - basically wrapping all strings with _(..) or gettext_noop(...) where appropriate. I also took the opportunity to make some of the messages more consistent with each other. (To cut down size, I'm not including the actual libvirt.pot file ) Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 5

[Libvir] problem with virConnectClose in libvirt 0.1.5
by Philippe Berthault 21 Sep '06

21 Sep '06

After closing libvirt with virConnectClose, it's possible to open libvirt again in the same program with virConnectOpen but following operations on domains systematically fail. In example: conn = virConnectOpen(NULL); dom = virDomainLookupByID(conn, id); virDomainFree(dom); virConnectClose(conn); ... conn = virConnectOpen(NULL); dom = virDomainLookupByID(conn, id); ==> ERROR The second virDomainLookupByID call fails because of the do_connect() function in xend_internal.c source file. In this function, the connect() API returns ENOENT on the second virDomainLookupByID() call. ENOENT return code isn't documented in the man of connect API ! I suppose that some cleaning are missing when virConnectClose is called or in the opposite, too cleaning is done. Regards.

2 2

[Libvir] New style CDROM config in Xen 3.0.3
by Daniel P. Berrange 14 Sep '06

14 Sep '06

So in a recent change to xen-unstable.hg the way CDROM devices are configured for HVM guests changed. They are now treated as regular IDE devices, and a CDROM is signified by presence of ':cdrom' suffix on the device name http://lists.xensource.com/archives/html/xen-devel/2006-08/msg00369.html libvirt obviously needs to support both this new style config, and the old (<= 3.0.2) style configs. Thus I requested addition of 'xend_config_format' attribute in the /xend/node SEXPR data from XenD. Christian recently committed this patch, so we can now support both CDROM style configs in libvirt. I am attaching a patch to implement this support - basically it touches xend_internal.c to fix XML->SEXPR conversion, and xml.c to fix SEXPR->XML conversion. Then this is corresponding additions to the unit tests to validate the new code, and also validate that I didn't break existing code. I am attaching the patches, but I don't intend to commit them to libvirt just yet because I don't actually have the new 3.0.3 xend running on my test systems fully yet. If anyone is using 3.0.3 any testing would be appreciated, otherwise I aim to test this towards end of next week. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

1 2

[Libvir] Configuration and building libvirt.
by Michel Gauthier 13 Sep '06

13 Sep '06

We want to configure and build libvirt library (0.1.4 for instance) from tar.gz file on a IA32 platform with RHEL4U2 distro without installing and build all Xen modules. The steps are the following: tar -xzvf libvirt-0.1.4.tar.gz cd libvirt-0.1.4 ./configure --prefix=/usr make make install Unfortunatly, the "./configure --prefix=/usr " aborts with: checking for library containing xs_read... no configure: error: Xen store library not found (see attached file) corresponding to lines: dnl search for the Xen store library AC_SEARCH_LIBS(xs_read, [xenstore], [], [AC_MSG_ERROR([Xen store library not found])]) in the configure file. We have then transfered the libxenstore.so file into /usr/lib directory of our machine, but the result is the same. (NB: We had previously transferred all Xen public C header files into /usr/include/xen directory) 1- Could you tell us what are the Xen files strictly necessary to achieve this operation? 2- Is this operation also possible with a IA64 building environment on our IA32 platform? Regards.

3 2

[Libvir] [PATCH] Support setting vnc port
by Jeremy Katz 12 Sep '06

12 Sep '06

We display the vncport, but never used it when creating the Xen sexpr. Attached fixes that. It also adds support for "port=-1" implying "use the first unused port"[1] Jeremy [1] The xen and qemu "vncunused" parameter

2 2

[Libvir] Proposal for virtual devices
by michel.ponceau＠bull.net 07 Sep '06

07 Sep '06

I previously made a proposal for block devices. It was rejected because it was VBD-specific and Xen-specific. Here is a proposal for transparent interface with Xen or other. The virtual device attributes (VDA) are transferred from/to application without considering their meaning, except backend domain, previously requested to be the name for user, while it is the ID for Xen. typedef struct virVDA { /* virtual device attribute */ char name[16]; char value[64]; } virVDA; int virDomainGetVdevs(virDomainPtr domain, char *type, virVDA *attr, int maxattr); * @type: pointer to device type, "disk", "network" ... * @attr: pointer to an array of virVDA structures * @maxattr: number of structures in attr array * Return: (number of VDAs per device)<<16 + (number of devices), if success; * -1 if failure. For "disk" type, Xen provides the following VDA names: "virtual-device" (virtual device number) "backend-id" (backend domain ID) "state" "backend" (backend device or file) "ring-ref" "event-channel" This function changes "backend-id" to "bkdom-name" and translates value ID into name. int virDomainCreateVdev(virDomainPtr domain, char *type, virVDA *attr, int nbattr); * @type: pointer to device type, "disk", "network" ... * @attr: pointer to an array of virVDA structures * @nbattr: number of structures in attr array For "disk" type, Xen requires the following VDA names: "uname" (backend device, value either "phy:<>" or "file:<>" "dev" (virtual device number) "mode" (value "ro" for read-only, or "rw" for read/write) "backend" (backend domain ID) This function translates "backend" user value (name) into ID. int virDomainDestroyVdev(virDomainPtr domain, char *type, char *ref); * @type: pointer to device type, "disk", "network" ... * @ref: pointer to virtual device reference (number for Xen...)

2 1

[Libvir] libvirt-0.1.5 release
by Daniel Veillard 05 Sep '06

05 Sep '06

With the recent Xen hypervisor breakage, I think it was urgent to push a new release for testing even if some parts are still in development and not finished. So expect a new release in a couple of weeks. In the meantime 0.1.5 is available at: ftp://libvirt.org/libvirt/ * Support for new hypercalls change in Xen changeset 86d26e6ec89b * bug fixes: - virParseUUID() was wrong, netwoking for paravirt guests (Daniel Berrange) - virsh on non-existent domains (Daniel Berrange) - string cast bug when handling error in python (Pete Vetere) - HTTP 500 xend error code handling (Pete Vetere and Daniel Berrange) * improvements: - test suite for SEXPR <-> XML format conversions (Daniel Berrange) - virsh output regression suite (Daniel Berrange) - new environ variable VIRSH_DEFAULT_CONNECT_URI for the default URI when connecting (Daniel Berrange) - graphical console support for paravirt guests (Jeremy Katz) - parsing of simple Xen config files (with Daniel Berrange) - early work on defined (not running) domains (Daniel Berrange) - virsh output improvement (Daniel Berrange) Thanks for everyone who helped out, please check that everything is working with the new hypervisor code, I could not test all entry points, I am on the road ! Daniel -- Red Hat Virtualization group http://redhat.com/virtualization/ Daniel Veillard | virtualization library http://libvirt.org/ veillard(a)redhat.com | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/

1 0

[Libvir] [Patch 3/4]: Xen inactive domains: stop xend/xs/xen using inactive domains
by Daniel P. Berrange 04 Sep '06

04 Sep '06

The existing xend, xs, xen, proxy driver backends all assume that if they're given a virDomainPtr object, the handle member is non-zero (ie its a running domain). With the addition of inactive domain support this assumption is no longer value. Thus this patch modifies all these drivers, so that if the handle is -1, then they return without doing any work. This lets the operation be correctly delegated to the new driver for inactive domains. In addition, the XenD driver assumed that if XenD returns a 404 when looking up a domain based on its name, this was a fatal error condition. Since inative domains are not known to XenD, this assumption is also now false. The attached patch modifies xend_internal so that if a 404 is returned, the error is not propagated to the libvirt error handlers. Instead the methods in question simply return -1, and control continues to the new backend for managing inactive domains. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 2

[Libvir] [Patch 2/4]: Xen inactive domains: hash file
by Daniel P. Berrange 04 Sep '06

04 Sep '06

The management of inactive domains uses the hashtable APIs for managing its list of config files. In doing so it needed some extra capabilities. This patch adds 3 new routines: - virHashForEach - iterate over all values in the hashtable invoking a callback for each one. Modifying the hashtable from the callback is forbidden, or bad stuff will happen. - virHashRemoteSet - iterate over all values in the hashtable invoking a callback for each one. If the callback returns non zero, the element will be removed. - virHashSearch - iterate over all values in the hashtable invoking a callback for each one. The first entry for which the callback returnbs non-zero will be returned. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 2

[Libvir] [Patch 1/4] Xen inactive domains: config file APIs
by Daniel P. Berrange 04 Sep '06

04 Sep '06

This patch adds an extra API to the config file management API. This API lets you add a new config parameter to the file. Previously you could only modify existing values, not define extra ones. It also exposes an API for creating a new empty config file record, which is needed if you're creating a brand new config file, rathre than processing an existing one. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 2

[Libvir] [Pach 4/4]: Xen inactive domains: new XM driver
by Daniel P. Berrange 04 Sep '06

04 Sep '06

This patch adds a new driver xm_internal.c which manages inactive domains whose config files are found in /etc/xen. On certain operations the driver will scan /etc/xen for new domains - the scanning is throttled to only occur at most every 10 seconds. When scanning any new domains are loaded into memory, and any domains deleted off disk are removed. This driver can start new domains, define domains, get/set memory, max memory, vcpus, can dump XML, and list inactive domains. The list of inactive domains is filtered to remove any domains which are running. This driver is still work in progress. There is more work needed to report error exact conditions via libvirt error APIs, and the code for serializing from XML -> XM config file is not yet complete. The code for serializing from XM config files into XML is pretty much complete although needs more testing. All the APIs can be tested via the commands added to virsh for inactive domains last week. eg list --inactive, start, define, undefine I'll post another update of this patch tomorrow which should be more complete Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

1 0

[Libvir] [Patch 0/4] Xen inactive domain
by Daniel P. Berrange 04 Sep '06

04 Sep '06

The following series of patches add support for managing inactive domains with Xen. Based on DV's recent xm config file reader, it processes all domains in /etc/xen (skipping known bogus files, xmexmaple*, xend-config.sxp). It can do all the operations you'd expect for managing inactive domains. Create a new domain, start a domain, list domains, delete a domain, change cpu / memory, get XML for domain. The impact on existing driver backends is very small - the only place which needed changing is the XenD code which lookups existing domains - making it be quiet if it get s a 404 error from XenD - this is to be expected if the domain is inactive. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

1 0

[Libvir] Semantics for ListDomains/ ListDefinedDomains
by Daniel P. Berrange 03 Sep '06

03 Sep '06

So I was thinking a little about lifecycle support for passive domains in libvirt and wanted to clarify the intended semantics of the two methods: virConnectListDomains virConnectListDefinedDomains Am I correct in thinking that virConnectListDefinedDomains will list a domain if-and-only-if its state == 'shutoff', and that virConnectListDomains will list a domain if-and-only-if its state != 'shutoff' ie, when I start a defined domain with 'virDomainCreate', it will be removed from the 'ListDefinedDomains' set and added to the 'ListDomains' set. Likewise when shutting down it'll move back the other way. Or will the virConnectListDefinedDomains call always list a domain with a config file, even if it is running ? I realize this is a little hypothetical since XenD doesn't have lifecycle management yet, but it matters to the test backend, and any potential QEMU / UML backend, and the future XenD XML-RPC backend Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

4 9

[Libvir] Libvirt / Xen incompatible versions.
by Michel Gauthier 01 Sep '06

01 Sep '06

We've just installed the last advanced version of Xen on our IA64 platform. Unfortunatly, the build of the 0.1.4 version of libvirt with this version of Xen is not successful. The C header file of Xen /usr/include/xen/dom0_ops.h used by libvirt in xen_internal.c has been modified. The dom0_getdomaininfo_t structure is no more decared in this header file. Are you aware of this problem? What is the latest version of Xen on which libvirt has been successfully built and tested? Thanks.

2 1

[Libvir] [QEMU 3/3] the qemu driver & daemon
by Daniel P. Berrange 31 Aug '06

31 Aug '06

Attached is the patch for the qemu driver & daemon process Since there is no scalable way to determine active standalone QEMU instances, or access their monitor connection, the backend implements a management daemon - called 'qemud' - which takes care of managing the complete lifecycle of QEMU virtual machines. The libvirt driver is a thin shim which talks to qemud via a UNIX domain socket - the wire protocol is modelled after the protocol used for libvirt_proxy, but obviously contains alot more functions since its full read/write, not just read-only. There is intended to be one qemud instance per UNIX user - the so called 'session' instance, and optionally one system global instance runing as root (or a system daemon account?) - the so called 'system' instance. When using libvirt, the URIs for each are 'qemu:///session' and 'qemu:///system' respectively. The UNIX socket for session instances is $HOME/.qemud and is chmod 0700 - no support is provided for users connecting to each other's instances. Although we could trivially extend to create a read-only socket $HOME/.qemud-ro if desired. The system instance is currently not finished but intended to run in /var/run/qemud or some such. The backend code is in src/qemu_internal.h, src/qemu_internal.c, the daemon code is in qemud/* with qemud/protocol.h defining the network wire protocol. This latter header file is also included by src/qemu_internal.c - qemud.c - the main daemon management code - handles all network I/O and forking / cleanup of QEMU processes. The implementation is poll() based & completely non-blocking. All requests/replies from a single client are, however, serialized. - dispatch.c - takes care of serializing/de-serializing requests/replies from/to libvirt - called from qemud.c when a complete message is available. - driver.c - the implementation of each request from libvirt - called from dispatch.c - config.c - manages reading & writing of XML files for inactive domains in the $HOME/.qemud.d/*.xml. Also takes care of generating a suitable command line for execing qemu. In terms of hardware, there is support for - harddisk - floppy, cdrom, harddisk - graphics - VNC - memory - virtual CPUs In particular no support for - networking - sound - non-native CPU types - boot order But its enough to demonstrate the practice. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

3 5

[Libvir] Added gettext translation support
by Daniel P. Berrange 31 Aug '06

31 Aug '06

The attached patch adds support for translating all text strings in libvirt and virsh using gettext. It also fixes up many error messages, and updates others to use a consistent style - so we don't have lots of slightly differenet translations of the same thing. It also adds an 'autoclean.sh' which deletes all autotools generated junk so you can more easily generate useful patches using diff against a pristine source tree. The patch adds a placeholder en_GB.po file (with no changes). It is expected that real translations for a large number of languages will be completed over course of next couple of months Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

1 0

[Libvir] [QEMU 2/3] add commands for inactive domains to virsh
by Daniel P. Berrange 30 Aug '06

30 Aug '06

The attached patch adds new commands to virsh, and modifies the list command. * 'list' is given two additional parameters '--inactive' to tell it to list inactive instances, instead of active instances, and --all to tell it to list both active & instance domains. NB inactive domains will show with an ID of -1 and state 'shut off'. * 'define' defines an inactive domain - instructing the backend to save the XML definition in some persist location. The domain is not started. * 'undefine' removes an inactive domain - the persist config file managed by the backend is deleted. The domain must not be running when this is run. * 'start' starts up an inactive domain which was previously defined via the 'define' command. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

3 5

[Libvir] [QEMU 1/3] hook up driver methods for inactive domains
by Daniel P. Berrange 30 Aug '06

30 Aug '06

The attached patch prepares the driver method API to support inactive domains. The following methods are hooked up: * virConnectListDefinedDomains * virConnectNumOfDefinedDomains * virDomainDefineXML * virDomainUndefine * virDomainCreate It sets all existing drivers to have NULL in these fields. The existing code in virDomainDefineXML was ripped out since it was found to be more trouble than it was worth - the backend driver needed to repeat most of the logic anyway - so better just to delegate the whole lot. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 3

[Libvir] generator.py -- why?
by pvetere＠redhat.com 29 Aug '06

29 Aug '06

Hi all. So, I found a bug in the python bindings that I'd really like to fix, but when I sat down to do so I quickly found myself mired in a swampy mess of code generation: generator.py. :-) Now, I feel compelled to ask -- why don't we just have a static libvirt.py file that is WYSIWYG? The generator.py program alone is longer than the file it generates, and having a static file would not only make the code easier to read, but would also make bug fixing much simpler. But, I'm sure there's got to be a good reason for it. :-) Here's a program that produces the bug I tried to address: import libvirt def get_domain(dom_name): conn = libvirt.openReadOnly(None) domain = conn.lookupByName(dom_name) return domain d = get_domain("mydomain") print d.info() This is a fairly typical "factory" pattern I could imagine people using. The problem is that the connection object falls out of scope after the get_domain routine ends, and is therefore garbage collected. This leaves the "d" domain object with no valid connection, resulting in a "libvir: Xen error : failed Xen syscall ioctl 3166208" error. The simple solution would be for the connection object to pass a reference to itself into each domain object it creates. The domain objects would then maintain the reference until they are destroyed. I valiantly tried to follow the table-based patterns used in generator.py to fix this problem, but I found myself inventing somewhat arbitrary rules just to force the code I wanted it to output. And hence, $SUBJECT. :-) Pete

4 7

[Libvir] [QEMU 0/3] QEMU support for libvirt
by Daniel P. Berrange 28 Aug '06

28 Aug '06

The following series of patches add support for QEMU to libvirt. The patches are: - libvirt-drivers-qemu.patch - hook up methods for inactive domains - libvirt-virsh-qemu.patch - hook up inactive domain methods to virsh - libvirt-backend-qemu.patch - the main QEMU backend The patches will be explained further in each email which follows. NB, the first two patches are ready for inclusion, but the last one needs more error handling, and another audit of the networking code to check in particular buffer sizes. For testing I used Damn Small Linux, and QEMU 0.8.1 from Fedora Core 5 (the one with VNC support) wget http://distro.ibiblio.org/pub/linux/distributions/damnsmall/current/dsl-3.0… Here is a suitable config file: <domain type='qemu'> <name>dsl</name> <uuid>fff64701-45b9-11cb-88e3-afbfe5370493</uuid> <memory>131072</memory> <vcpu>20</vcpu> <devices> <disk type='file' device='cdrom'> <source file='/home/berrange/dsl-3.0.1.iso'/> <target dev='hdc'/> </disk> <graphics type='vnc' port='5906'/> </devices> </domain> An example session with virsh: [berrange@aubergine ~]$ export VIRSH_DEFAULT_CONNECT_URI=qemu:///session [berrange@aubergine ~]$ virsh list Id Name State ---------------------------------- [berrange@aubergine ~]$ virsh define $HOME/dsl.xml Domain dsl defined from /home/berrange/dsl.xml [berrange@aubergine ~]$ virsh list --all Id Name State ---------------------------------- -1 dsl shut off [berrange@aubergine ~]$ virsh dominfo dsl Id: -1 Name: dsl UUID: fff64701-45b9-11cb-88e3-afbfe5370493 OS Type: linux State: shut off CPU(s): 20 Max memory: 131072 kB Used memory: 131072 kB [berrange@aubergine ~]$ virsh start dsl Domain dsl started [berrange@aubergine ~]$ virsh list --all Id Name State ---------------------------------- 1 dsl running [berrange@aubergine ~]$ virsh dumpxml dsl <domain type='qemu' id='1'> <name>dsl</name> <uuid>fff64701-45b9-11cb-88e3-afbfe5370493</uuid> <memory>131072</memory> <vcpu>20</vcpu> <devices> <disk type='file' device='cdrom'> <source file='/home/berrange/dsl-3.0.1.iso'/> <target dev='hdc'/> <readonly/> </disk> <graphics type='vnc' port='5906'/> </devices> </domain> [berrange@aubergine ~]$ virsh destroy dsl Domain dsl destroyed [berrange@aubergine ~]$ virsh list --all Id Name State ---------------------------------- -1 dsl shut off I'm also happy to report that it is also possible to use virt-manager with the QEMU support - just enter qemu:///session as the URI. That said you can't use the VM creation wizard, but it should be easy to hook that up. The fact that virt-manager works without modification is a really great proof of the value of libvirt abstract driver backend API concept - validating the decision not to make a Xen specific API. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

1 0

[Libvir] Support graphical console for PV Xen guests
by Jeremy Katz 26 Aug '06

26 Aug '06

With the paravirt framebuffer patches for Xen, we can support having a graphical console for PV as well as HVM guests. The attached adds support to libvirt by breaking out the <graphics> parsing into its own function that can be used by both PV and HVM guests. It also adds support for looking at the appropriate sexpr for the vnc/sdl console for PV guests. One thing that is potentially not okay is that I've changed sexpr_node to take a format string instead of just a const char *, but since it's not public API and doesn't actually change anything for callers I think it should be fine. If not, let me know and I can change it. Jeremy

2 3

[Libvir] RFC: A more convenient 'virsh create' command
by Daniel P. Berrange 26 Aug '06

26 Aug '06

The current implementation of 'virsh create' takes an XML file as its only parameter & creates a domain from this. This is great if you have a suitable XML file already, but if you are just trying to automate some simple tasks from the shell then the need to use XML is a little cumbersome. Thus I was thinking perhaps we could have an alternate way to define a new VM (keep the current XML based way too of course) QEMU for example makes it very easy to launch a new VM: qemu -m 256 -hda /path/to/image.dsk -hdc /path/to/boot.iso Taking inspiration from this syntax we could allow: virsh start -m 256 -hda /path/to/image.dsk -hdc /path/to/boot.iso -name Foo Internally, the 'start' command would simply transform these command line args into the neccessary libvirt XML and then call the normal create functions. Another way would be have a 'genxml' command, which accepted these list of devices / config properties & then printed out appropriate XMl. This could be piped to the regular 'virsh create' command virsh genxml -m 256 -hda /path/to/image.dsk -hdc /path/to/boot.iso \ -name Foo | virsh create - This isn't so critical for Xen, because people are already used to writing config files before creating the domain, but when we get a QEMU backend i think such a convenient method for defining new VMs will be neccessary to encourage users to use virsh instead of manually calling 'qemu'. Even for Xen users it would make shell script easier though :-) Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

3 5