June 2009 - Devel - libvirt List Archives

[libvirt] [PATCH] Fix python error reporting for some storage operations
by Cole Robinson 07 Mar '25

07 Mar '25

In the python bindings, all vir* classes expect to be passed a virConnect object when instantiated. Before the storage stuff, these classes were only instantiated in virConnect methods, so the generator is hardcoded to pass 'self' as the connection instance to these classes. Problem is there are some methods that return pool or vol instances which aren't called from virConnect: you can lookup a storage volume's associated pool, and can lookup volumes from a pool. In these cases passing 'self' doesn't give the vir* instance a connection, so when it comes time to raise an exception crap hits the fan. Rather than rework the generator to accomodate this edge case, I just fixed the init functions for virStorage* to pull the associated connection out of the passed value if it's not a virConnect instance. Thanks, Cole diff --git a/python/generator.py b/python/generator.py index 01a17da..c706b19 100755 --- a/python/generator.py +++ b/python/generator.py @@ -962,8 +962,12 @@ def buildWrappers(): list = reference_keepers[classname] for ref in list: classes.write(" self.%s = None\n" % ref[1]) - if classname in [ "virDomain", "virNetwork", "virStoragePool", "virStorageVol" ]: + if classname in [ "virDomain", "virNetwork" ]: classes.write(" self._conn = conn\n") + elif classname in [ "virStorageVol", "virStoragePool" ]: + classes.write(" self._conn = conn\n" + \ + " if not isinstance(conn, virConnect):\n" + \ + " self._conn = conn._conn\n") classes.write(" if _obj != None:self._o = _obj;return\n") classes.write(" self._o = None\n\n"); destruct=None

4 3

[libvirt] [PATCH 0/5] Add virConnectListDefinedInterfaces/NumOfDefinedInterfaces
by Laine Stump 16 Jul '09

16 Jul '09

This patch adds everything but the backend implementation of two APIs that were requested during review of my patch to add virsh commands that expose the virInterface* API. These count and list interfaces on the host that are currently inactive (ie "down"). virConnectListInterfaces and virConnectNumOfInterfaces will list/count *only* those interfaces that are currently active. With these functions in place, I'll be able to add the "--inactive" and "--all" flags to iface-list, as suggested by danpb. Implementation of the backend of these two functions is waiting for two things: 1) supporting code in libnetcf, and 2) implementation of the rest of the backend driver functions for virInterface* (ready in the wings and waiting to go).

3 14

[libvirt] More then one pci device possible with libvirt and pci passthrough?
by Mirko Raasch 09 Jul '09

09 Jul '09

Hi, i want to use more then one pci card in my debian guest. But when i start the guest, the log says: libvirtd: 15:44:55.459: warning : pciTrySecondaryBusReset:483 : Other devices on bus with 0000:05:01.0, not doing bus reset With "/usr/bin/qemu-system-x86_64 -m 512 -name test -hda /media/vm/test.img -boot c -vnc :2 -k de -pcidevice host=05:01.0 -pcidevice host=05:02.0 -pcidevice host=05:03.0" there seems to be no problems. How can i start my guest with three pci devices passed through and libvirt? My hostdev section in the xml looks like this: <hostdev mode='subsystem' type='pci' managed='yes'> <source> <address domain='0x0000' bus='0x05' slot='0x01' function='0x0'/> </source> </hostdev> <hostdev mode='subsystem' type='pci' managed='yes'> <source> <address domain='0x0000' bus='0x05' slot='0x02' function='0x0'/> </source> </hostdev> <hostdev mode='subsystem' type='pci' managed='yes'> <source> <address domain='0x0000' bus='0x05' slot='0x03' function='0x0'/> </source> </hostdev> Greetings Mirko

3 5

[libvirt] kernel summit topic - 'containers end-game'
by Serge E. Hallyn 08 Jul '09

08 Jul '09

A topic on ksummit agenda is 'containers end-game and how do we get there'. So for starters, looking just at application (and system) containers, what do the libvirt and liblxc projects want to see in kernel support that is currently missing? Are there specific things that should be done soon to make containers more useful and usable? More generally, the topic raises the question... what 'end-games' are there? A few I can think of off-hand include: 1. resource control 2. lightweight virtual servers 3. (or 2.5) unprivileged containers/jail-on-steroids (lightweight virtual servers in which you might, just maybe, almost, be able to give away a root account, at least as much as you could do so with a kvm/qemu/xen partition) 4. checkpoint, restart, and migration For each end-game, what kernel pieces do we think are missing? For instance, people seem agreed that resource control needs io control :) Containers imo need a user namespace. I think there are quite a few network namespace exploiters who require sysfs directory tagging (or some equivalent) to allow us to migrate physical devices into network namespaces. And checkpoint/restart needs... checkpoint/restart. thanks, -serge

5 14

[libvirt] [PATCH] Power Hypervisor Support for libvirt - minimum set of features
by Eduardo Otubo 07 Jul '09

07 Jul '09

Hello all, This is the initial patch for the driver for IBM Power Hypervisors. The minimum set of features are now implemented: list, list --all and dumpxml. Here is the Changeset since last PATCH I sent: * The URI has changed to: phyp://user@[hmc|ivm]/managed_system. If the system is a HMC+VIOS based, only an HMC authentication will be required. Commands will be sent to VIOS trough HMC command line. And if the system is an IVM based, then just provide the username and password for IVM. * Since the Power Hypervisor has no information about UUID's, I built a little database (uuid_db) to store and associate LPAR ID's with UUID randomly generated by the API. * The command dumpxml is implemented, but there are some informations missing. Fetching informations like fstab, os type, uptime, IP addr and so on, will only be available in a future versions of the HMC system. * The TODO list is now set to implement life cycle functions. Thanks in advance, []'s -- Eduardo Otubo Software Engineer Linux Technology Center IBM Systems & Technology Group Mobile: +55 19 8135 0885 otubo(a)linux.vnet.ibm.com

2 7

[libvirt] PATCH: 0/3: Run QEMU guests within a CGroup
by Daniel P. Berrange 07 Jul '09

07 Jul '09

Recent Linux kernels have a new concept of 'CGroups' which is a way to group tasks on the system and apply policy to them as a whole. We already use this in the LXC container driver, to control total memory usage of things runing within a container. This patch series is a proof of concept to make use of CGroups in the QEMU driver. The idea is that we have a 3 level cgroup hierarchy - Top level; contains the libvirtd daemon itself - 2nd level: one per libvirt driver, but dos not contain any processes. - 3rd level: one per guest VM. Contains the QEMU process The host admin can do control on the top level and 2nd level to set an overall system policy. libvirt will then provide APIs / capabilities to control individual VMs policy. Daniel -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

4 10

[libvirt] [PATCH] Do a correct format mapping of partitions 2
by Henrik Persson 06 Jul '09

06 Jul '09

I make a new try from a different mail client now. /Henrik

4 5

[libvirt] vTPM support in libvirt
by Andreas Sommer 02 Jul '09

02 Jul '09

I'm wondering if there will be vTPM support in libvirt in the near future?! Xen does support it already with the configuration "vtpm = ['instance=1,backend=0']", for example. So it would be great if the libvirt XML format supported it, too... For example like this: <devices> <vtpm instance="1" backend="xxx" /> </devices> Both attributes are optional. The backend attribute is a VM ID (on Xen, only zero for dom0 is supported) and could be implemented as a UUID?! What do you think? Are there any efforts to introducing that? Cheers Andreas

6 13

[libvirt] [Patch] Fix network driver not to forget to save <domain/> in network XML file
by Satoru SATOH 02 Jul '09

02 Jul '09

It seems that libvirt network driver can understand /network/domain element but it does not save this configuration into the network definition XML file correctly (looks completely forgetting about it). Here is a fix for this problem. (see also: rhbz#505720) --- src/network_conf.c | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/src/network_conf.c b/src/network_conf.c index 1e0cbb8..1b50b84 100644 --- a/src/network_conf.c +++ b/src/network_conf.c @@ -592,6 +592,9 @@ char *virNetworkDefFormat(virConnectPtr conn, def->stp ? "on" : "off", def->delay); + if (def->domain) + virBufferVSprintf(&buf, " <domain name='%s'/>\n", def->domain); + if (def->ipAddress || def->netmask) { virBufferAddLit(&buf, " <ip"); -- 1.6.3.2

3 2

[libvirt] [Patch] Network schema fixes and updates
by Satoru SATOH 02 Jul '09

02 Jul '09

This is a patch to make network schema (network.rng) to support some neat features (/netowrk/domain, //dhcp/host/*). --- docs/schemas/network.rng | 14 ++++++++++++++ 1 files changed, 14 insertions(+), 0 deletions(-) diff --git a/docs/schemas/network.rng b/docs/schemas/network.rng index 34c0163..a4281a5 100644 --- a/docs/schemas/network.rng +++ b/docs/schemas/network.rng @@ -1,6 +1,7 @@  <element name="network" xmlns="http://relaxng.org/ns/structure/1.0" datatypeLibrary="http://www.w3.org/2001/XMLSchema-datatypes"> + <interleave>  <element name="name"><text/></element> @@ -43,6 +44,11 @@ </element> </optional> <optional> + <element name="domain"> + <attribute name="name"><text/></attribute> + </element> + </optional> + <optional>  <!-- FIXME: address, netmask and the start and end of the ranges @@ -63,7 +69,15 @@ <attribute name="end"><text/></attribute> </element> </zeroOrMore> + <zeroOrMore> + <element name="host"> + <attribute name="mac"><text/></attribute> + <attribute name="name"><text/></attribute> + <attribute name="ip"><text/></attribute> + </element> + </zeroOrMore> </element> </element> </optional> + </interleave> </element> -- 1.6.3.2

3 2

[libvirt] [PATCH 0/1] Fix pool create when pool already exists
by David Allan 02 Jul '09

02 Jul '09

When storagePoolCreate is called to define a pool that already exists, it mistakenly destroys the existing pool. This patch fixes the problem. Dave

5 8

[libvirt] PATCH : Opennebula, adding xen i686 domain capability and updating some error messages
by "Abel Míguez Rodríguez" 01 Jul '09

01 Jul '09

Hi, this patch uses "OPERATION_INVALID" error instead "OPERATION_FAILED" error where should be used, also adds a missing capability, defining Xen i686 domains. Thanks for the patches, --- one_driver.c.bak 2009-06-30 15:02:28.000000000 +0200 +++ one_driver.c 2009-06-30 15:02:29.000000000 +0200 @@ -505,7 +505,7 @@ static int oneDomainShutdown(virDomainPt ret= 0; goto return_point; } - oneError(dom->conn, dom, VIR_ERR_OPERATION_FAILED, + oneError(dom->conn, dom, VIR_ERR_OPERATION_INVALID, _("Wrong state to perform action")); goto return_point; } @@ -541,7 +541,7 @@ static int oneDomainDestroy(virDomainPtr if(c_oneCancel(vm->pid)) { /* VM not running, delete the instance at ONE DB */ if(c_oneFinalize(vm->pid)){ - oneError(dom->conn, dom, VIR_ERR_OPERATION_FAILED, + oneError(dom->conn, dom, VIR_ERR_OPERATION_INVALID, _("Wrong state to perform action")); goto return_point; } @@ -576,11 +576,11 @@ static int oneDomainSuspend(virDomainPtr ret=0; goto return_point; } - oneError(dom->conn, dom, VIR_ERR_OPERATION_FAILED, + oneError(dom->conn, dom, VIR_ERR_OPERATION_INVALID, _("Wrong state to perform action")); goto return_point; } - oneError(dom->conn,dom,VIR_ERR_OPERATION_FAILED, + oneError(dom->conn,dom,VIR_ERR_OPERATION_INVALID, _("domain is not running")); } else { oneError(dom->conn, dom, VIR_ERR_INVALID_DOMAIN, @@ -609,11 +609,11 @@ static int oneDomainResume(virDomainPtr ret=0; goto return_point; } - oneError(dom->conn, dom, VIR_ERR_OPERATION_FAILED, + oneError(dom->conn, dom, VIR_ERR_OPERATION_INVALID, _("Wrong state to perform action")); goto return_point; } - oneError(dom->conn,dom,VIR_ERR_OPERATION_FAILED, + oneError(dom->conn,dom,VIR_ERR_OPERATION_INVALID, _("domain is not paused ")); } else { oneError(dom->conn, dom, VIR_ERR_INVALID_DOMAIN, --- one_conf.c.bak 2009-06-30 15:01:06.000000000 +0200 +++ one_conf.c 2009-06-30 15:00:58.000000000 +0200 @@ -94,6 +94,27 @@ virCapsPtr oneCapsInit(void) { goto no_memory; } + if ((guest = virCapabilitiesAddGuest(caps, + "xen", + "i686", + 32, + NULL, + NULL, + 0, + NULL)) == NULL) + { + goto no_memory; + } + if (virCapabilitiesAddGuestDomain(guest, + "one", + NULL, + NULL, + 0, + NULL) == NULL) + { + goto no_memory; + } + return caps; Abel Miguez Rodriguez ---- Distributed System Architecture Group (http://dsa-research.org) GridWay, http://www.gridway.org OpenNEbula, http://www.opennebula.org

3 4

[libvirt] [PATCH] logging: remove unused level 0
by Amy Griffis 01 Jul '09

01 Jul '09

Some more logging cleanup. Remove all references to log level 0, and replace with VIR_LOG_DEBUG where appropriate. Also change any hard numbers to the corresponding VirLogPriority and add VIR_LOG_DEFAULT for the starting log level. Signed-off-by: Amy Griffis <amy.griffis(a)hp.com --- docs/logging.html.in | 6 ++---- qemud/libvirtd.conf | 5 ++--- qemud/qemud.c | 11 ++++++----- src/logging.c | 10 +++++----- src/logging.h | 2 ++ 5 files changed, 17 insertions(+), 17 deletions(-) diff --git a/docs/logging.html.in b/docs/logging.html.in index 11ea68c..fcd100f 100644 --- a/docs/logging.html.in +++ b/docs/logging.html.in @@ -49,7 +49,6 @@ <ul> <li>log_level: accepts the following values: <ul> - <li>0: no logging at all</li> <li>4: only errors</li> <li>3: warnings and errors</li> <li>2: informations, warnings and errors</li> @@ -84,20 +83,19 @@ </ul> In all cases the x prefix is the minimal level, acting as a filter: <ul> - <li>0: everything</li> <li>1: DEBUG</li> <li>2: INFO</li> <li>3: WARNING</li> <li>4: ERROR</li> </ul> Multiple output can be defined , they just need to be separated by - spaces, e.g.: <code>"3:syslog:libvirtd 0:file:/tmp/libvirt.log"</code> + spaces, e.g.: <code>"3:syslog:libvirtd 1:file:/tmp/libvirt.log"</code> will log all warnings and errors to syslog under the libvirtd ident but also log everything debugging and informations included in the file <code>/tmp/libvirt.log</code> For example setting up the following: <pre>export LIBVIRT_DEBUG=1 -export LIBVIRT_LOG_OUTPUTS="0:file:virsh.log"</pre> +export LIBVIRT_LOG_OUTPUTS="1:file:virsh.log"</pre> and then running virsh will accumulate the logs in the <code>virsh.log</code> file in a way similar to: <pre>14:29:04.771: debug : virInitialize:278 : register drivers diff --git a/qemud/libvirtd.conf b/qemud/libvirtd.conf index 1fd5918..49de466 100644 --- a/qemud/libvirtd.conf +++ b/qemud/libvirtd.conf @@ -269,7 +269,7 @@ # Logging controls # -# Logging level: 0 none, 4 errors, 3 warnings, 2 informations, 1 debug +# Logging level: 4 errors, 3 warnings, 2 informations, 1 debug # basically 1 will log everything possible #log_level = 3 @@ -303,13 +303,12 @@ # x:file:file_path # output to a file, with the given filepath # In all case the x prefix is the minimal level, acting as a filter -# 0: everything # 1: DEBUG # 2: INFO # 3: WARNING # 4: ERROR # -# Multiple output can be defined , they just need to be separated by spaces. +# Multiple output can be defined, they just need to be separated by spaces. # e.g.: # log_outputs="3:syslog:libvirtd" # to log all warnings and errors to syslog under the libvirtd ident diff --git a/qemud/qemud.c b/qemud/qemud.c index b5e3665..da20aa9 100644 --- a/qemud/qemud.c +++ b/qemud/qemud.c @@ -128,7 +128,7 @@ static int sigwrite = -1; /* Signal handler pipe */ static int ipsock = 0; /* -l Listen for TCP/IP */ /* Defaults for logging */ -static int log_level = 3; +static int log_level = VIR_LOG_DEFAULT; static char *log_filters = NULL; static char *log_outputs = NULL; @@ -2499,15 +2499,16 @@ qemudSetLogging(virConfPtr conf, const char *filename) { log_outputs = strdup(debugEnv); } if (!log_outputs) { + char *tmp = NULL; if (godaemon) { - char *tmp = NULL; if (virAsprintf (&tmp, "%d:syslog:libvirtd", log_level) < 0) goto free_and_fail; - virLogParseOutputs (tmp); - VIR_FREE (tmp); } else { - virLogParseOutputs("0:stderr:libvirtd"); + if (virAsprintf(&tmp, "%d:stderr", log_level) < 0) + goto free_and_fail; } + virLogParseOutputs(tmp); + VIR_FREE(tmp); } else { virLogParseOutputs(log_outputs); } diff --git a/src/logging.c b/src/logging.c index fd1d5d4..d13ae0f 100644 --- a/src/logging.c +++ b/src/logging.c @@ -119,7 +119,7 @@ static int virLogNbOutputs = 0; /* * Default priorities */ -static virLogPriority virLogDefaultPriority = VIR_LOG_WARN; +static virLogPriority virLogDefaultPriority = VIR_LOG_DEFAULT; static int virLogResetFilters(void); static int virLogResetOutputs(void); @@ -174,7 +174,7 @@ int virLogStartup(void) { virLogLen = 0; virLogStart = 0; virLogEnd = 0; - virLogDefaultPriority = VIR_LOG_WARN; + virLogDefaultPriority = VIR_LOG_DEFAULT; virLogUnlock(); return(0); } @@ -196,7 +196,7 @@ int virLogReset(void) { virLogLen = 0; virLogStart = 0; virLogEnd = 0; - virLogDefaultPriority = VIR_LOG_WARN; + virLogDefaultPriority = VIR_LOG_DEFAULT; virLogUnlock(); return(0); } @@ -689,7 +689,7 @@ int virLogParseOutputs(const char *outputs) { virSkipSpaces(&cur); while (*cur != 0) { prio= virParseNumber(&cur); - if ((prio < 0) || (prio > 4)) + if ((prio < VIR_LOG_DEBUG) || (prio > VIR_LOG_ERROR)) return(-1); if (*cur != ':') return(-1); @@ -770,7 +770,7 @@ int virLogParseFilters(const char *filters) { virSkipSpaces(&cur); while (*cur != 0) { prio= virParseNumber(&cur); - if ((prio < 0) || (prio > 4)) + if ((prio < VIR_LOG_DEBUG) || (prio > VIR_LOG_ERROR)) return(-1); if (*cur != ':') return(-1); diff --git a/src/logging.h b/src/logging.h index d99bb31..0b9ae79 100644 --- a/src/logging.h +++ b/src/logging.h @@ -77,6 +77,8 @@ typedef enum { VIR_LOG_ERROR, } virLogPriority; +#define VIR_LOG_DEFAULT VIR_LOG_WARN + /** * virLogOutputFunc: * @category: the category for the message

3 3

[libvirt] feature or bug?
by Gene Czarcinski 30 Jun '09

30 Jun '09

Yet another question. I am using the virt-manager GUI. When I begin to define a new guest and attempt to "browse" to specify where the iso image file is, the GUI dialog brings me up in my home directory. But, almost everything associated with libvirt is done as the root user. By default, why isn't the initial directory "/var/lib/libvirt/images"? Gene

2 1

[libvirt] your patence is appreciated
by Gene Czarcinski 30 Jun '09

30 Jun '09

I really hate being a bit of a PITA with all of my question/comments, but I am just trying to use Fedora virtualization in a manner where I can reasonably manage the configuration. I suspect I am not the first ex-VMware user and I will not be the last such user who is trying to use Fedora virtualization in place of VMware. I am very familiar with how guests are "managed" with VMware. I want Fedora (and Linux) virtualization to be a good product. Your patience is appreciated! Gene

2 1

[libvirt] need some help/asdive
by Gene Czarcinski 30 Jun '09

30 Jun '09

Background: I am a newbie to Linux Virtualization. However, I have been using Linux for a long time. After a short time with Slackware, I started with Red Hat Linux 2 and went on to 2.1, 3.0.3, 4, 4.2, 5, 5.2, ..., 8, 9. And then on to Fedora Core 1, 2, ... until I installed Fedora 11 about a month ago. I have also run VMware starting when it first appeared. Initially I used VMware to run Windows. These days I use it to run up to from 25 to 30 guests (between 1 and 5 at the same time) [yes, I have lots of system memory]. Besides a couple of Windows guests, I have some guests supporting both 32 bit (i386) and 64 bit (x86_64) architectures of older releases as well as alternatives such as CentOS -- I use these to build RPMs. Then there are the guests supporting other operating systems such as Solaris and *BSD. Finally, I have a set of test guest that I use for testing stuff which could destroy the operating system or which require a lot of re-booting. All of this in 25 to 30 guests. With Fedora 11, I started by trying VMware-server. Server 1.0.9 (which I used on Fedora 9) looks like much more work to get running ... more or less "a bridge too far". So, I tried server 2.0.2 ... but the client interface changed from a separate GUI application to a browser based interface: 1) this impacts my usage of the firefox browser; 2) it is a buggy implementation; and 3) it is a crappy interface which I do not like. So, I (finally) noticed that Linux in general and Fedora specifically has packages supporting running guests virtually ... so, lets change my "problem sets" and give this a try ... especially when I noticed that it could use the hardware support in my new AMD Phenom II 940 processor. Before getting into my specific questions/problem, let me describe how my systems are set up. First of all, I have a basic rule: "don't screw with a working system." Therefore, on each of my systems I have at least two root partitions defined -- one for the current (working) system and one for the next (new) system [these days it is actually a /boot partition with root being on a striped logical volume]. I also have a small (think minimal install) partition with Fedora to serve as a boot selector (using chainloader), as something to to define striped logical volumes, and to be there for "emergency" maintenance. Guys, this works for me and always doing fresh installs saves on other headaches. Implicit in the above is that I put all of my data and my /home into separate partitions or logical volumes. One of these are my VMware guest systems. Now VMware puts all of the files (configuration, disk image, etc.) associated a guest under a separate directory. Fedora virtualization does not do this and separation is something I need so that I can easily move to a new version of the operating system such as when Fedora 12 becomes available in November. Questions/problems: My first question: it appears to me that a guest is defined by two things -- a configuration file in /etc/libvirt/qemu and one or more disk images in /var/lib/libvirt/images ... is this it or are there other files I need to consider? My second question concerns the location of the disk images: what is the "best practical way" to move these out of the root partition? I have come up with these different approaches which could/may work: 1. change the "path" value in /etc/libvirt/storage/default.xml to point to a directory in a different partition. 2. Use a /etc/fstab entry to "bind mount" a directory in a separate partition onto /var/lib/libvirt/images 3. Just use an /etc/fstab entry to mount a partition on /var/lib/libvirt/images or, perhaps, on /var/lib/libvirt. [This could be done during system installation] OK, any comments on these?? I can live with changing some configuration files, I can live with copying guest configuration files between systems, I CANNOT live with the disk image files being put into the root partition. Before jumping "whole hog" into using Fedora Virtualization in place of VMware, I need to be able to live with the result. Figure 30 guest virtuals with 10GB of virtual disk each, that is 300GB. I have that space but not for every root partition on my system! Comment: It sure would be nice if Fedora Virtualization put all files related to a guest under a single directory the way VMware does. Comment: It sure would be nice if there was some system-wide configuration parameter that defined where virtualization files went. Right now, as a newbie, I am using the virt-manage GUI interface. I am studying the command-line programs to see what they offer me. So far, I like what I see in Fedora Virtualization but I need to address how things will be practically managed sooner rather than later. Any help appreciated. Gene

2 4

[libvirt] PATCH: Remove unused Open Nebula code, fix potential missing NULL terminator
by Daniel P. Berrange 30 Jun '09

30 Jun '09

This patch removes a bunch of unused Open Nebula code which has a number of flaws. It also fixes handling of a strncpy() call to ensure there is a NULL terminator in the copied string. And finally it makes a couple of strings be 'const' since they're never changed Daniel Index: src/opennebula/one_client.c =================================================================== RCS file: /data/cvs/libvirt/src/opennebula/one_client.c,v retrieving revision 1.1 diff -u -p -r1.1 one_client.c --- src/opennebula/one_client.c 24 Jun 2009 17:32:56 -0000 1.1 +++ src/opennebula/one_client.c 30 Jun 2009 11:05:12 -0000 @@ -33,16 +33,9 @@ void c_oneStart() xmlrpc_client_init2(&one_client.env, XMLRPC_CLIENT_NO_FLAGS, "OpenNebula API Client", "1.2", NULL, 0); - one_client.error=0; - -#ifdef ONED_PORT - one_client.url=(char *)malloc(64); - snprintf(one_client.url, 63, "http://localhost:%d/RPC2", ONED_PORT); -#else - one_client.url=(char *)"http://localhost:2633/RPC2"; -#endif - - one_client.session=(char *)"one-session"; + one_client.error = 0; + one_client.url = "http://localhost:2633/RPC2"; + one_client.session = "one-session"; }; @@ -94,35 +87,6 @@ int c_oneMigrate(int vmid, int hid, int return c_oneReturnCode(resultP); } -int c_oneAllocate(char* template_file) -{ - int file; - int size; - int bytes_read; - char *file_text; - int return_code; - - file=open(template_file, O_RDONLY); - size=lseek(file, 0, SEEK_END); - lseek(file, 0, SEEK_SET); - - file_text=(char *)malloc(size+1); - bytes_read=read(file, file_text, size); - close(file); - - if(bytes_read==size) { - file_text[size]=0; - - return_code=c_oneAllocateTemplate(file_text); - free(file_text); - - return return_code; - } else { - free(file_text); - return -1; - } -} - int c_oneAllocateTemplate(char* vm_template) { xmlrpc_value *resultP; @@ -222,6 +186,7 @@ int c_oneVmInfo(int vmid, char* ret_info if( return_code ) { strncpy(ret_info, return_string, length-1); + ret_info[length-1] = '\0'; xmlrpc_DECREF(resultP); free(return_string); @@ -241,9 +206,6 @@ void c_oneFree() { xmlrpc_env_clean(&one_client.env); xmlrpc_client_cleanup(); -#ifdef ONE_PORT - free(one_client.url); -#endif } Index: src/opennebula/one_client.h =================================================================== RCS file: /data/cvs/libvirt/src/opennebula/one_client.h,v retrieving revision 1.1 diff -u -p -r1.1 one_client.h --- src/opennebula/one_client.h 24 Jun 2009 17:32:56 -0000 1.1 +++ src/opennebula/one_client.h 30 Jun 2009 11:05:12 -0000 @@ -24,8 +24,8 @@ struct _oneClient { xmlrpc_env env; - char *url; - char *session; + const char *url; + const char *session; char *error; }; @@ -38,8 +38,6 @@ int c_oneDeploy(int vmid, int hid); int c_oneMigrate(int vmid, int hid, int flag); -int c_oneAllocate(char* template_file); - int c_oneAllocateTemplate(char* vm_template); int c_oneAction(int vmid,char* action); -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

2 1

[libvirt] problem: deleting a guest
by Gene Czarcinski 30 Jun '09

30 Jun '09

I defined a new guest using the virt-manager GUI ... everything fine so far. I then decided that my definition was not correct (or I could be done with that guest) and could not find a way to delete the guest. OK, quit virt-manager, delete the configuration and disk-image files for the guest, restart libvirtd, and then start virt-manager again. The guest definition (according to virt-manager) was still there. I finnaly got rid of the definition by re-booting the system. THIS IS NOT SATISFACTORY! IMHO, I should be able to: 1. Define a new guest. 2. Delete the definition of a new guest but not the related configuration and disk-image files. 3. Delete the guest definition and the related configuration and disk image files but NOT and CD/DVD image files external to /var/lib/libvirt/images. 4. Import new guests which were created as clones, copied from another host, or convert created from something like VMware. Not it is certainly possible (or even likely) that I h\just do not understand how to do things yet. Or, this could be a bug. Or, this could be a lack of capabilities. Help appreciated in pointing me in the right direction. I will bugzilla where appropriate. Gene

3 3

[libvirt] How does libvirt 0.6.3 load kvm driver
by Jason Wong 30 Jun '09

30 Jun '09

Dear all, I found that libvirt 0.6.3 seems does not load KVM driver while libvirt 0.3.3 does. Situation description: I am using CentOS 5.3, and the kernel I used is 2.6.30. I downloaded the kernel from kernel.org andcompile it. KVM was compiled as a module. libvirt: libvirt 0.6.3, released by Eucalyptus(a cloud computing infrastructure software which supports KVM as well) (1) libvirt 0.6.3 does not load KVM driver After booting the machine with kernel 2.6.30 and probe KVM related module. I started libvirtd 0.6.3 and set LD_LIBRARY_PATH to "$LIBVIRT_063_HOME/lib:$LIBVIRT_063_HOME/usr/lib:$LD_LIBRARY_PATH". I found that I am not able to connect to libvirtd using "virsh --connect qemu:///system"(the virsh used is virsh 0.6.3 that shipped with libvirt 0.6.3), here is the debug message: ======================================= debug message ========================================= 17:41:11.219: debug : virInitialize:290 : register drivers 17:41:11.219: debug : virRegisterDriver:667 : registering Test as driver 0 17:41:11.219: debug : virRegisterNetworkDriver:567 : registering Test as network driver 0 17:41:11.219: debug : virRegisterStorageDriver:598 : registering Test as storage driver 0 17:41:11.219: debug : virRegisterDeviceMonitor:629 : registering Test as device driver 0 17:41:11.219: debug : virRegisterDriver:667 : registering Xen as driver 1 17:41:11.219: debug : vboxRegister:68 : VBoxCGlueInit failed: 0.0.0, errorval=-1 17:41:11.219: debug : virRegisterDriver:667 : registering VBOX as driver 2 17:41:11.219: debug : virConnectOpen:1050 : name=qemu:///system 17:41:11.219: debug : do_open:920 : name "qemu:///system" to URI components: scheme qemu opaque (null) authority (null) server (null) user (null) port 0 path /system 17:41:11.219: debug : do_open:930 : trying driver 0 (Test) ... 17:41:11.219: debug : do_open:936 : driver 0 Test returned DECLINED 17:41:11.219: debug : do_open:930 : trying driver 1 (Xen) ... 17:41:11.219: debug : do_open:936 : driver 1 Xen returned DECLINED 17:41:11.219: debug : do_open:930 : trying driver 2 (VBOX) ... 17:41:11.219: debug : do_open:936 : driver 2 VBOX returned DECLINED libvir: error : could not connect to qemu:///system 17:41:11.219: debug : virUnrefConnect:210 : unref connection 0x60dd00 1 17:41:11.219: debug : virReleaseConnect:171 : release connection 0x60dd00 libvirt error: could not connect to qemu:///system (code=5) error: failed to connect to hypervisor ================================end of debug message ========================================= It seems that libvirt 0.6.3 does not load QEMU driver. (2) libvirt 0.3.3 does load KVM driver However, when I run the command "lsmod | grep kvm". I found KVM module was present: ================================= result of lsmod ============================================== [root@node2 ~]# lsmod | grep kvm kvm_intel 46664 0 kvm 166224 1 kvm_intel [root@node2 ~]# ================================= end of result of lsmod ========================================= I stopped libvirtd 0.6.3 and then started libvirtd 0.3.3(the one shipped with CentOS 5.3). The virtual machine manager works well and I can use "virsh --connect qemu:///system"(the virsh used is virsh 0.3.3 that shipped with CentOS 5.3) to connect to libvirtd. I dig into libvirt source code. I found that in libvirt 0.3.3: There is qemudRegister() in libvirt.c. I think this function call is responsible for loading qemu(kvm) driver. And in libvirt 0.6.3: I cannot found qemudRegister() or virDriverLoadModule("qemu") in the source code. My question is: Does libvirt 0.6.3 support KVM? If yes, how does it load KVM driver? I want to find out why libvirt 0.3.3 works well while libvirt 0.6.3 does not work in my computer. Thanks. -- Jason Wong System engineer Cluster Technology Limited Email: hcwong(a)clustertech.com Direct Line: (852) 2655-6129 Tel: (852) 2655-6100 Fax: (852) 2994-2101 Website: www.clustertech.com ------------------------------ ************************************************************************** The information and attachment contained in this e-mail originating from Cluster Technology Limited is confidential and intended solely for the specified addressees. If you have received this email in error, please do not read, copy, distribute, disclose or use any information of this email in any way and please immediately notify the sender and delete this email. Thank you for your cooperation. **************************************************************************

2 4

[libvirt] hooking a physical nic to a bridge with net-create
by Shahar Klein 29 Jun '09

29 Jun '09

Hi I trying to set up a network configuration using libvirt The setup is: 1. a VM using virtio using a tap device connected to a bridge device 2. on the host side eth0 added to the same bridge device(without ip) using libvirt application (say virsh) I want to launch this set up: virsh net-create for the second part virsh create for the first part(tap, virtio, vm etc) I can set the first part(vm, tap, virtio) using the domain xml but I'm also interested in XMLing the second part and net-create it. I tried but only suceeded in creating the bridge Is it possible? am I missing somthing? Thanks in advance

4 7

[libvirt] HACKING file on website
by Daniel P. Berrange 29 Jun '09

29 Jun '09

FYI, I just committed a patch to add the contents of the 'HACKING' file to the website, under the section 'Documentation -> Internals -> Contributor Guidelines No content changes from HACKING, just html-ized existing content. Regards, Daniel -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

1 0

[libvirt] identi.ca micro-blogging group for libvirt & related apps
by Daniel P. Berrange 29 Jun '09

29 Jun '09

Micro-blogging is very popular these days and I know a number of people working on / with libvirt are using twitter/identi.ca So I decided it was worth creating a libvirt group on identi.ca for targetting notices relating to libvirt and any apps built using libvirt. You can find it here http://identi.ca/group/libvirt Any posting you make containing the magic word '!libvirt' will appear in the group page. Regards, Daniel -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

1 0

[libvirt] [PATCH] Fixing issues in re-detection of transient VMs
by Federico Simoncelli 29 Jun '09

29 Jun '09

Re-detection of transient VMs had two issues described in bugs 507304 and 507537: https://bugzilla.redhat.com/show_bug.cgi?id=507304 Summary: After a migration the domain status is saved as "paused" and it is not updated to "running". A following libvirtd restart will detect the vm as "paused" (instead of "running"). https://bugzilla.redhat.com/show_bug.cgi?id=507537 Summary: Destroying a transient re-detected domain leaves the domain defined (instead of completely remove it). Patches in attachment (and available in the bug report). -- Federico Simoncelli.

2 2

[libvirt] [PATCH] don't break on libnuma failures
by Daniel Veillard 28 Jun '09

28 Jun '09

Another patch hanging in bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=506590 sometimes libnuma can't handle some topologies, but those failures should not completely break libvirt, patch by Dan Berrange. Look sane to me, fixes the main problem for the reporter, so I applied and commited it, 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] unchecked malloc+syscall uses
by Jim Meyering 28 Jun '09

28 Jun '09

I was looking through recent changes and spotted two potential NULL-dereferences in this change set: http://git.et.redhat.com/?p=libvirt.git;a=commitdiff;h=70c01b4c1adea75e9 In src/opennebula/one_client.c: + one_client.url=(char *)malloc(64); + snprintf(one_client.url, 63, "http://localhost:%d/RPC2", ONED_PORT); +#else + one_client.url=(char *)"http://localhost:2633/RPC2"; +#endif -- + file_text=(char *)malloc(size+1); + bytes_read=read(file, file_text, size); + close(file); + + if(bytes_read==size) { Perhaps of more immediate concern, these syscalls are not checked for failure: + file=open(template_file, O_RDONLY); + size=lseek(file, 0, SEEK_END); + lseek(file, 0, SEEK_SET); and same for the "read" syscall above. Note, this is just the result of a few minutes cursory review.

2 1