- Devel - libvirt List Archives

Réf. : Re: [Libvir] Re: Proposal : add 3 functions to Libvirt API, for virtual CPUs
by jean-paul.pigache＠bull.net 21 Jul '06

21 Jul '06

This interesting thread is temporarilu suspended because Michel has left for some (long) vacation until August 15th. I think the last proposal is a good one and deserves to be implemented and tested. I am confident that Michel will be willing to update its patch accordingly when he comes back. Is it OK for you ?. Side question: is there a defined "sign-off" procedure to follow when submitting patches ? Cordialement. Jean-Paul Daniel Veillard <veillard(a)redhat.com> Envoyé par : libvir-list-bounces(a)redhat.com 17/07/2006 15:08 Veuillez répondre à veillard Pour : "Daniel P. Berrange" <berrange(a)redhat.com> cc : libvir-list(a)redhat.com, michel.ponceau(a)bull.net Objet : Re: [Libvir] Re: Proposal : add 3 functions to Libvirt API, for virtual CPUs On Mon, Jul 17, 2006 at 01:50:25PM +0100, Daniel P. Berrange wrote: > On Tue, Jul 11, 2006 at 10:50:53AM -0400, Daniel Veillard wrote: > > We still have a relatively simple API for the common case, and for special > > cases we have an extension capability with relatively clear definitions. it's > > a bitstrange but I think that should cover most case as best as possible > > I dont particularly like this as an API because I think it will be error > prone for application developers. Most app developers will only ever have > a handful of CPUs in their test machines, so they'll never the alternate > codepath for > 256 cpu case. Likewise I don't like the idea of a virVcpuInfo > struct which has a variable size because it will totally confuse people who > haven't read the API docs very carefully, again leading to obscure bugs. > > The root problem is that we have two conflicting goals here > > 1. Want to have virVcpuInfo be a fixed size struct > 2. We want a cpumap of arbitrary size Right this is confusing, I wanted to avoid another allocation in the general case but this made things even more confusing. > The obvious solution to this problem is to *remove* the cpumap data from > the virVcpuInfo structure completely, and always pass in a separately > malloc'd array of the correct size. So I'd suggest: > > typedef struct _virVcpuInfo virVcpuInfo; > struct _virVcpuInfo { > unsigned int number; /* virtual CPU number */ > int state; /* value from virVcpuState */ > unsigned long long cpuTime; /* CPU time used, in nanoseconds */ > int cpu; /* real CPU number, or -1 if offline */ > } > > virDomainGetVcpus(virDomainPtr domain, virVcpuInfoPtr info, int maxinfo, > char *cpumap, int maplen); Agreed, this is the most logical approach > > The client applications calling this API already have to malloc() the memory > region for the 'info' parameter of a correct size, so having to also malloc > the cpumap parameter is no extra trouble. > > virDomainInfo info; > virDomainVpuInfoPtr cpuInfo; > int cpuMapLen; > char *cpuMap; > > virDomainGetInfo(domain, &info); > > cpuInfo = malloc(sizeof(virDomainVcpuInfo)*info.nrVirtCpu); > cpuMapLen = (info.nrVirtCpu + 7) / 8 ; > cpuMap = malloc(cpuMapLen); > > virDomainGetVCpus(domain, cpuInfo, info.nrVirtCpu, cpuMap, cpuMapLen); > > ... do stuff with the data ... > > free(cpuInfo); > free(cpuMap); > > > So you can see there is minimal extra work to always pass in cpuMap as > a separate parameter. If an application didn't care about the cpuMap > data they could simply pass in NULL. Agree. We should keep an example of use (like above completed) in the source tree to help developpers. Daniel -- Daniel Veillard | Red Hat http://redhat.com/ veillard(a)redhat.com | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/ -- Libvir-list mailing list Libvir-list(a)redhat.com https://www.redhat.com/mailman/listinfo/libvir-list

2 1

[Libvir] Bug with libvirt in Xen 3.0.1?
by pvetere＠redhat.com 21 Jul '06

21 Jul '06

Hi, I think I may have found a bug in libvirt and wanted see what people thought. I'm using the stock FC5 installation at the moment (with xen 3.0.1), and the newest version of libvirt. I am noticing that with xen 3.0.1 and newer versions of libvirt, getDomainsID() seems to return bogus values. For example: [root@test05 ~]# virsh list Id Name State ---------------------------------- 0 Domain-0 running libvir: Xen Daemon error : GET operation failed: No such domain 65486 libvir: Xen Daemon error : GET operation failed: No such domain 2986 But, if I use xm, I get what I expect: [root@test05 ~]# xm list Name ID Mem(MiB) VCPUs State Time(s) Domain-0 0 1507 2 r----- 387.5 vm1 1 256 1 -b---- 151.3 vm2 2 256 1 -b---- 87.9 After some digging around in the code, I believe that libvirt is incorrectly identifying the hypervisor as being "old" in xen_internal.c:xenHypervisorInit, and is therefore passing the incorrect parameter structure into the hypervisor when it makes its ioctl in xen_internal.c:xenHypervisorListDomains. I've tried this same test on a system running xen 3.0.2, and as I expected everything works fine. So, there must be something different about xen 3.0.1 that libvirt is not accounting for. At this point, I don't really have more time to dig further but I thought I'd bring up the issue in case someone on this list can offer more insight. Pete

3 5

[Libvir] Virtual Machine Manager 0.1.5 release
by Daniel P. Berrange 20 Jul '06

20 Jul '06

I've just pushed out another snapshot of the Virtual Machine Manager GUI application for managing Xen machines via libvirt. A summary of the changes in this 0.1.5 release[1]: * Integration with GNOME keyring for the VNC console to avoid need to remember passwords when accessing the guest domain's console * Use cairo to render a '50% alpha gray wash' over the screenshot to give appearance of a 'dimmed' display when paused. Also render the word 'paused' in big letters. * Initial cut of code for saving & restoring domain snapshots to disk * Added icons for buttons which were missing graphics * Miscellaneous RPM spec file fixes to comply with rpmlint * Update status icons to match those in the gnome-applet-vm * Added domain ID and # VCPUs to summary view * Misc bug fixes The next priority target for development is some UI for creating new VM instances (a GUI equivalent of xenguest-install.py script shipping with current Fedora Xen packages) and implenenting the UI for editing domain configuration / hardware. In addition there is now a more permanent home for the application's web pages: http://virt-manager.et.redhat.com/index.html And the Mercurial repository is now published for anonymous read-only access at: http://hg.et.redhat.com/virt/applications/virt-manager--devel Patches are welcome either as traditional unified diffs, or as mercurial changeset bundles - see the following page for details on exporting bundles using hg: http://virt-manager.et.redhat.com/scmrepo.html Regards, Dan. [1] There was a 0.1.4 but it had a fatal flaw in it breaking the console -- |=- 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] Question about XML
by Juan Walker 20 Jul '06

20 Jul '06

Hi, I have a question regarding the XML format that libvir uses for domain configuration. I was wondering if there is a way to use the configuration format that xm uses or if a tool exists to convert to XML. I'm writing an app to help users manage their Xen domains but I'm hoping to be able to use any .sxp files that the user already has. I suspect I missed something in the documentation about why the XML format is better, but if there is no such tool maybe I should volunteer to write one? Thanks.

3 2

[Libvir] just a question about libvirt evolution
by Philippe Berthault 20 Jul '06

20 Jul '06

Hello, Several major modifications will appear this year in future Xen releases (3.0.3 and 3.0.4) that may have impacts on libvirt. These modifications (see xen roadmap edited by Ian Pratt in july 2006) are : - Introduction of XML-RPC, - Xen configuration files in XML format, - Suppression of S-Expressions, - etc.. What is the libvirt position about these points ? Is the initial goal of libvirt to mask Xen evolutions will be always possible ? The most problematic problem would be the configurations files in XML format. Libvirt already supports its own XML format of configuration files. It seems to me that it would be interesting to use the same XML format (or a compatible format) in libvirt and in Xen. Is there a plan in this direction ? Thanks.

3 2

[Libvir] Patch to correct spelling
by Daniel P. Berrange 20 Jul '06

20 Jul '06

The attached patch changes 'informations' to 'information' throughout the codebase to fix up the English grammer. I assume its ok to commit ? 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] possible bug in virDomainLookupByID
by Bryan D. Payne 19 Jul '06

19 Jul '06

I'm experiencing an odd problem in virDomainLookupByID when using 0.1.3 that does not exist when using 0.1.1. Perhaps this is a side effect of switching the backend connection? I'm curious is anyone else has seen this and/or if you have any suggestions for fixing it. The attached code generates the problem that I'm seeing. Specifically, after upgrading to 0.1.3, calling virDomainLookupByID twice results in a failure on the second call. This behavior was not seen in 0.1.1. Any thoughts? Thanks, bryan PS... This problem is seen in XenAccess. Thanks to Daniele Sgandurra for bringing it to my attention. #include <stdio.h> #include <string.h> #include <libvirt/libvirt.h> #include <libvirt/virterror.h> #include <libxml/parser.h> #include <libxml/tree.h> char *get_xml_info (int id) { virConnectPtr conn = NULL; virDomainPtr dom = NULL; char *xml_data = NULL; /* NULL means connect to local Xen hypervisor */ conn = virConnectOpenReadOnly(NULL); if (NULL == conn) { printf("ERROR: Failed to connect to hypervisor\n"); goto error_exit; } /* Find the domain of the given id */ dom = virDomainLookupByID(conn, id); if (NULL == dom) { virErrorPtr error = virConnGetLastError(conn); printf("(%d) %s\n", error->code, error->message); printf("ERROR: Failed to find Domain %d\n", id); goto error_exit; } xml_data = virDomainGetXMLDesc(dom, 0); error_exit: if (NULL!= dom) virDomainFree(dom); if (NULL!= conn) virConnectClose(conn); return xml_data; } int main () { char *xml = NULL; xml = get_xml_info(1); if (!xml){ printf("Failed on first try\n"); return 0; } free(xml); xml = get_xml_info(1); if (!xml){ printf("Failed on second try\n"); return 0; } free(xml); return 1; }

3 5

[Libvir] virDomainLookupByID problem when Root
by Daniele Sgandurra 17 Jul '06

17 Jul '06

Hi all! I've tried to compile and execute info1.c both as user and as root. As user I have no problem, but as root I get: Failed to find Domain 0 I've traced back the problem to the function: virDomainLookupByID(virConnectPtr conn, int id) at the line: path = xs_get_domain_path(conn->xshandle, (unsigned int) id); When root I get path != NULL, otherwise path==NULL; if path==NULL then the block after it is executed and everything is ok, otherwise at the end of the function name=null. (I manually changed path to NULL so that I can run it when root and it works even if info.maxMem it's different from when the program is run as a normal user: it shows -4 instead of 464896). It could be a configuration problem of mine (of Xen Store for example)? Thanks! Chiacchiera con i tuoi amici in tempo reale! http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com

3 3

[Libvir] announce: gnome-applet-vm-0.1.0rc1
by Karel Zak 14 Jul '06

14 Jul '06

I have just pushed out a new version of the VM Applet. The changes/features are: - D-BUS support and integration with virt-manager - console helper support for privileged operations - fixed a lot of bugs - new icons - better GUI - the applet internally supports multiple libvirt connection (in future it will be useful for monitoring remote virtual machines) Web page (with screenshot of course:-): http://people.redhat.com/kzak/gnome-applet-vm/ Please, help me found bugs and improve this application. Karel -- Karel Zak <kzak(a)redhat.com>

1 0

[Libvir] Patch: Fix for UUIDs in XML parser
by Peter Vetere 13 Jul '06

13 Jul '06

Hi all. Here's a pretty straightforward patch that fixes a bug I found where, if <uuid>...</uuid> is specified in the XML on the call createLinux, it is not propagated down to Xend. The result is an instance with a different UUID than the one requested. Pete

2 1