- Devel - libvirt List Archives

[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

[libvirt] More then one pci device possible with libvirt and pci passthrough?
by Mirko Raasch 28 Jun '09

28 Jun '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

1 0

[libvirt] Can't scale XP on remote
by sean darcy 27 Jun '09

27 Jun '09

Just starting out with an XP guest on Fedora 11 kvm. I'm using virt-manager to connect from a laptop. Server: kernel-2.6.29.4-167.fc11.x86_64 qemu-*-0.10.50-6.kvm86.fc11.x86_64 libvirt-0.6.4-2.fc11.x86_64 On the laptop: virt-manager-0.7.0-5.fc11.i586 libvirt-0.6.2-11.fc11.i586 The console comes, and XP runs fine. But, screen is scaled too large vertically for my laptop screen which is 1280x800. Only with fullscreen can I see the bottom menu bar. I would have assumed I could change the size of the screen (even if not the aspect ratio), but I can't. Am I doing something wrong, or is this just The Way Things Are? sean

3 6

[libvirt] [PATCH] avoid dumpXML segfault with recent Xen (3.4.0)
by Daniel Veillard 26 Jun '09

26 Jun '09

This is basically the patch by Sascha at https://bugzilla.redhat.com/show_bug.cgi?id=503254 with a couple of cleanups. I think it's low risk, and avoid a crash so I commited it. Basically ths SXPR syntax changed from (type vnc) to (vnc 1) for VNC vfb devices, and assuming something similar for sdl. The code checks first for type then backoff for vnc/sdl so I think this is backward compatible. 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] [PATCH 3/5] disk geometry
by Henrik Persson 26 Jun '09

26 Jun '09

Reads the disk geometry to be able to align disk partitions on cylinder boundries. Msdos partition tables really like this stuff.

4 3

[libvirt] [PATCH 0/5] Create storage volumes with disk backend
by Henrik Persson 26 Jun '09

26 Jun '09

Hi, These patches make it possible to create storage volumes when having a disk as backend. As sent in earlier mail, it does not handle LVM and RAID partition currently. /Henrik

2 1

[libvirt] [PATCH] Don't disable all logging with --disable-debug
by Amy Griffis 26 Jun '09

26 Jun '09

This patch changes the automake conditional for debugging to disable only the debug output, not the entire logging module. Since the logging symbols are always visible now, I moved them to libvirt_private.syms and got rid of libvirt_debug.syms. I noticed that debugFlag wasn't being used anymore, so I removed that too. There was some code in __virExec where if you called with NULLs for childout and childerr, it would conditionally set the fds to either -1 or the /dev/null fd depending on ENABLE_DEBUG. I'm not sure why the difference, but it seemed safe to remove the entire conditional section since either setting is effectively treated the same way in the following code. Signed-off-by: Amy Griffis <amy.griffis(a)hp.com> diff --git a/src/Makefile.am b/src/Makefile.am index f65e7ad..23a1843 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -498,10 +498,6 @@ EXTRA_DIST += \ # symbols aren't present at all. # -if ENABLE_DEBUG -USED_SYM_FILES += libvirt_debug.syms -endif - if WITH_DRIVER_MODULES USED_SYM_FILES += libvirt_driver_modules.syms endif @@ -517,7 +513,6 @@ endif EXTRA_DIST += \ libvirt_public.syms \ libvirt_private.syms \ - libvirt_debug.syms \ libvirt_driver_modules.syms \ libvirt_bridge.syms \ libvirt_linux.syms diff --git a/src/libvirt.c b/src/libvirt.c index bf49018..bcb707e 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -256,9 +256,7 @@ winsock_init (void) int virInitialize(void) { -#ifdef ENABLE_DEBUG char *debugEnv; -#endif if (initialized) return(0); @@ -269,7 +267,6 @@ virInitialize(void) virRandomInitialize(time(NULL) ^ getpid())) return -1; -#ifdef ENABLE_DEBUG debugEnv = getenv("LIBVIRT_DEBUG"); if (debugEnv && *debugEnv && *debugEnv != '0') { if (STREQ(debugEnv, "2") || STREQ(debugEnv, "info")) @@ -287,7 +284,6 @@ virInitialize(void) debugEnv = getenv("LIBVIRT_LOG_OUTPUTS"); if (debugEnv) virLogParseOutputs(debugEnv); -#endif DEBUG0("register drivers"); @@ -1055,13 +1051,11 @@ do_open (const char *name, /* Secondary driver for storage. Optional */ for (i = 0; i < virStorageDriverTabCount; i++) { res = virStorageDriverTab[i]->open (ret, auth, flags); -#ifdef ENABLE_DEBUG DEBUG("storage driver %d %s returned %s", i, virStorageDriverTab[i]->name, res == VIR_DRV_OPEN_SUCCESS ? "SUCCESS" : (res == VIR_DRV_OPEN_DECLINED ? "DECLINED" : (res == VIR_DRV_OPEN_ERROR ? "ERROR" : "unknown status"))); -#endif if (res == VIR_DRV_OPEN_ERROR) { if (0 && STREQ(virStorageDriverTab[i]->name, "remote")) { virLibConnWarning (NULL, VIR_WAR_NO_STORAGE, diff --git a/src/libvirt_debug.syms b/src/libvirt_debug.syms deleted file mode 100644 index 1742a0b..0000000 --- a/src/libvirt_debug.syms +++ /dev/null @@ -1,19 +0,0 @@ -# -# These symbols are dependent upon --enable-debug. -# - - -# libvirt_internal.h -debugFlag; - - -# logging.h -virLogMessage; -virLogSetDefaultPriority; -virLogDefineFilter; -virLogDefineOutput; -virLogParseFilters; -virLogParseOutputs; -virLogStartup; -virLogShutdown; -virLogReset; diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 309d7f9..3958d9b 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -192,6 +192,18 @@ virRegisterStorageDriver; virRegisterDeviceMonitor; +# logging.h +virLogMessage; +virLogSetDefaultPriority; +virLogDefineFilter; +virLogDefineOutput; +virLogParseFilters; +virLogParseOutputs; +virLogStartup; +virLogShutdown; +virLogReset; + + # memory.h virAlloc; virAllocN; diff --git a/src/logging.c b/src/logging.c index 83e07cc..fd1d5d4 100644 --- a/src/logging.c +++ b/src/logging.c @@ -39,9 +39,6 @@ #include "util.h" #include "threads.h" -#ifdef ENABLE_DEBUG -int debugFlag = 0; - /* * Macro used to format the message as a string in virLogMessage * and borrowed from libxml2 (also used in virRaiseError) @@ -793,5 +790,4 @@ int virLogParseFilters(const char *filters) { } return(ret); } -#endif /* ENABLE_DEBUG */ diff --git a/src/logging.h b/src/logging.h index 7ea8935..d99bb31 100644 --- a/src/logging.h +++ b/src/logging.h @@ -32,22 +32,17 @@ #ifdef ENABLE_DEBUG #define VIR_DEBUG_INT(category, f, l, fmt,...) \ virLogMessage(category, VIR_LOG_DEBUG, f, l, 0, fmt, __VA_ARGS__) +#else +#define VIR_DEBUG_INT(category, f, l, fmt,...) \ + do { } while (0) +#endif /* !ENABLE_DEBUG */ + #define VIR_INFO_INT(category, f, l, fmt,...) \ virLogMessage(category, VIR_LOG_INFO, f, l, 0, fmt, __VA_ARGS__) #define VIR_WARN_INT(category, f, l, fmt,...) \ virLogMessage(category, VIR_LOG_WARN, f, l, 0, fmt, __VA_ARGS__) #define VIR_ERROR_INT(category, f, l, fmt,...) \ virLogMessage(category, VIR_LOG_ERROR, f, l, 0, fmt, __VA_ARGS__) -#else -#define VIR_DEBUG_INT(category, f, l, fmt,...) \ - do { } while (0) -#define VIR_INFO_INT(category, f, l, fmt,...) \ - do { } while (0) -#define VIR_WARN_INT(category, f, l, fmt,...) \ - do { } while (0) -#define VIR_ERROR_INT(category, f, l, fmt,...) \ - do { } while (0) -#endif /* !ENABLE_DEBUG */ #define VIR_DEBUG(fmt,...) \ VIR_DEBUG_INT("file." __FILE__, __func__, __LINE__, fmt, __VA_ARGS__) @@ -108,8 +103,6 @@ typedef int (*virLogOutputFunc) (const char *category, int priority, */ typedef void (*virLogCloseFunc) (void *data); -#ifdef ENABLE_DEBUG - extern int virLogSetDefaultPriority(int priority); extern int virLogDefineFilter(const char *match, int priority, int flags); extern int virLogDefineOutput(virLogOutputFunc f, virLogCloseFunc c, @@ -128,18 +121,4 @@ extern void virLogMessage(const char *category, int priority, const char *funcname, long long linenr, int flags, const char *fmt, ...) ATTRIBUTE_FORMAT(printf, 6, 7); -#else /* ENABLE_DEBUG */ - -#define virLogSetDefaultPriority(p) -#define virLogDefineFilter(m, p, f) -#define virLogDefineOutput(func, c, d, p, f) -#define virLogStartup() -#define virLogReset() -#define virLogShutdown() -#define virLogParseFilters(f) -#define virLogParseOutputs(o) -#define virLogMessage(c, p, func, l, f, fmt, __VA_ARGS__) - -#endif /* ENABLE_DEBUG */ - #endif diff --git a/src/util.c b/src/util.c index 8b746f5..408e17a 100644 --- a/src/util.c +++ b/src/util.c @@ -349,10 +349,6 @@ __virExec(virConnectPtr conn, } else { childout = *outfd; } -#ifndef ENABLE_DEBUG - } else { - childout = null; -#endif } if (errfd != NULL) { @@ -380,10 +376,6 @@ __virExec(virConnectPtr conn, } else { childerr = *errfd; } -#ifndef ENABLE_DEBUG - } else { - childerr = null; -#endif } if ((pid = fork()) < 0) {

3 2

[libvirt] [PATCH 0/1] Fix bogus WWNs
by David Allan 26 Jun '09

26 Jun '09

When HBA kernel modules are loaded while libvirt is running, libvirt will often read bogus WWNs from sysfs as at least some modules trigger the HAL event notifying libvirt of adapters' existence before the sysfs info is correct. The following patch causes libvirt to reread the WWNs whenever dumpxml is called so the capability information about the WWNs will be up to date. I pulled the logic to read the WWNs into two separate functions, and the patch is slightly more invasive as a result. The resulting code is smaller and cleaner for it, though, so I think this is the right way to go. (The diff is also very ugly; it should make sense once applied.) Dave

3 3

[libvirt] virStorageVolCreateXMLFrom
by Kenneth Nagin 26 Jun '09

26 Jun '09

I'd like to get a better understanding of the virStorageVolCreateXMLFrom api Its name and signature imply that it is a cloning operation that copies a volume from one location to another. Is this correct? Assuming that the answer is yes I have few other questions: Does it support remote copy of non-shared storage, e.g. from a http server? Is there any thought about adding digest validation check, e.g. with a md5 digest? Is copy-on-write cloning supported in this command? Kenneth Nagin

3 3

[libvirt] [PATCH 5/5] create storage volumes, take 2
by Henrik Persson 26 Jun '09

26 Jun '09

Here is a updated patch that defines a enum for mapping parition code to parted's fs-type. /Henrik

1 0

[libvirt] [PATCH] For comment/critique only - netcf backend for virInterface*.
by Laine Stump 26 Jun '09

26 Jun '09

This is the backend for the interface driver (virInterface*()) that uses netcf. There are a few issues with it: 1) It doesn't yet implement the backend for virConnectListDefinedInterfaces() and virConnectNumOfDefinedInterfaces() (although it does use my patched netcf API to list only active interfaces for virConnectListInterfaces()). (That patch isn't committed to netcf yet, btw; it's functional, but a bit rough and hackish) 2) interfaceLookupByMACString() still uses the old arg convention for ncf_lookup_by_mac_string(), which returned a single interface. That function has been changed to potentially return multiple interfaces. To fully support that, we would need to change the libvirt API. What should we do here? 3) The XML in interfaceDefineXML() isn't validated, although the spot to do it is clearly marked. 4) There's no comment in the place where we could, in the future, transform the XML returned to interfaceGetXMLDesc() by netcf. Currently the two use the same RNG. Since it would be a null transform, and that will only change under our control, that isn't an *immediate* problem, but shouldn't be forgotten. Parts that I'm unsure of: 1) changes to the Makefile, particularly surrounding "DRIVER_MODULES". 2) Is it registered in the proper places? 3) Could it (should it?) be added to other hypervisor drivers? (Currently I only put it in qemud.c) --- qemud/Makefile.am | 4 + qemud/qemud.c | 6 + src/Makefile.am | 16 ++- src/interface_driver.c | 381 ++++++++++++++++++++++++++++++++++++++++++++++++ src/interface_driver.h | 34 +++++ 5 files changed, 440 insertions(+), 1 deletions(-) create mode 100644 src/interface_driver.c create mode 100644 src/interface_driver.h diff --git a/qemud/Makefile.am b/qemud/Makefile.am index 403846a..9f982ba 100644 --- a/qemud/Makefile.am +++ b/qemud/Makefile.am @@ -134,6 +134,10 @@ if WITH_NETWORK libvirtd_LDADD += ../src/libvirt_driver_network.la endif +if WITH_NETCF + libvirtd_LDADD += ../src/libvirt_driver_interface.la +endif + if WITH_NODE_DEVICES libvirtd_LDADD += ../src/libvirt_driver_nodedev.la endif diff --git a/qemud/qemud.c b/qemud/qemud.c index b5e3665..9326ce9 100644 --- a/qemud/qemud.c +++ b/qemud/qemud.c @@ -81,6 +81,9 @@ #ifdef WITH_NETWORK #include "network_driver.h" #endif +#ifdef WITH_NETCF +#include "interface_driver.h" +#endif #ifdef WITH_STORAGE_DIR #include "storage_driver.h" #endif @@ -822,6 +825,9 @@ static struct qemud_server *qemudInitialize(int sigread) { #ifdef WITH_NETWORK networkRegister(); #endif +#ifdef WITH_NETCF + interfaceRegister(); +#endif #ifdef WITH_STORAGE_DIR storageRegister(); #endif diff --git a/src/Makefile.am b/src/Makefile.am index f65e7ad..f5bf1b8 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -154,6 +154,9 @@ ONE_DRIVER_SOURCES = \ NETWORK_DRIVER_SOURCES = \ network_driver.h network_driver.c +INTERFACE_DRIVER_SOURCES = \ + interface_driver.h interface_driver.c + # Storage backend specific impls STORAGE_DRIVER_SOURCES = \ storage_driver.h storage_driver.c \ @@ -381,8 +384,18 @@ libvirt_driver_network_la_SOURCES = $(NETWORK_DRIVER_SOURCES) endif if WITH_NETCF -libvirt_driver_interface_la_CFLAGS = $(NETCF_CFLAGS) libvirt_driver_interface_la_LDFLAGS = $(NETCF_LIBS) +libvirt_driver_interface_la_CFLAGS = $(NETCF_CFLAGS) +if WITH_DRIVER_MODULES +mod_LTLIBRARIES += libvirt_driver_interface.la +else +noinst_LTLIBRARIES += libvirt_driver_interface.la +libvirt_la_LIBADD += libvirt_driver_interface.la +endif +if WITH_DRIVER_MODULES +libvirt_driver_interface_la_LDFLAGS += -module -avoid-version +endif +libvirt_driver_interface_la_SOURCES = $(INTERFACE_DRIVER_SOURCES) endif # Needed to keep automake quiet about conditionals @@ -467,6 +480,7 @@ EXTRA_DIST += \ $(OPENVZ_DRIVER_SOURCES) \ $(VBOX_DRIVER_SOURCES) \ $(NETWORK_DRIVER_SOURCES) \ + $(INTERFACE_DRIVER_SOURCES) \ $(STORAGE_DRIVER_SOURCES) \ $(STORAGE_DRIVER_FS_SOURCES) \ $(STORAGE_DRIVER_LVM_SOURCES) \ diff --git a/src/interface_driver.c b/src/interface_driver.c new file mode 100644 index 0000000..b6cf510 --- /dev/null +++ b/src/interface_driver.c @@ -0,0 +1,381 @@ +/* + * interface_driver.c: backend driver methods to handle physical + * interface configuration using the netcf library. + * + * Copyright (C) 2006-2009 Red Hat, Inc. + * Copyright (C) 2006 Daniel P. Berrange + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Author: Daniel P. Berrange <berrange(a)redhat.com> + */ + +#include <config.h> + +#include <sys/types.h> +#include <sys/poll.h> +#include <dirent.h> +#include <limits.h> +#include <string.h> +#include <stdio.h> +#include <strings.h> +#include <stdarg.h> +#include <stdlib.h> +#include <unistd.h> +#include <errno.h> +#include <sys/utsname.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <signal.h> +#include <paths.h> +#include <pwd.h> +#include <stdio.h> +#include <sys/wait.h> +#include <sys/ioctl.h> +#include <netcf.h> + +#include "virterror_internal.h" +#include "datatypes.h" +#include "interface_driver.h" +#include "driver.h" +#include "event.h" +#include "buf.h" +#include "util.h" +#include "memory.h" +#include "logging.h" + +#define VIR_FROM_THIS VIR_FROM_INTERFACE + +#define interfaceReportError(conn, dom, net, code, fmt...) \ + virReportErrorHelper(conn, VIR_FROM_THIS, code, __FILE__, \ + __FUNCTION__, __LINE__, fmt) + +/* Main driver state */ +struct interface_driver +{ + virMutex lock; + struct netcf *netcf; +}; + + +static void interfaceDriverLock(struct interface_driver *driver) +{ + virMutexLock(&driver->lock); +} + +static void interfaceDriverUnlock(struct interface_driver *driver) +{ + virMutexUnlock(&driver->lock); +} + + +static struct interface_driver *driverState = NULL; + +static virDrvOpenStatus interfaceOpenInterface(virConnectPtr conn, + virConnectAuthPtr auth ATTRIBUTE_UNUSED, + int flags ATTRIBUTE_UNUSED) +{ + if (VIR_ALLOC(driverState) < 0) + { + virReportOOMError(conn); + goto alloc_error; + } + + /* initialize non-0 stuff in driverState */ + if (virMutexInit(&driverState->lock) < 0) + { + /* what error to report? */ + goto mutex_error; + } + + /* open netcf */ + if (ncf_init(&driverState->netcf, NULL) != 0) + { + /* what error to report? */ + goto netcf_error; + } + + conn->interfacePrivateData = driverState; + return 0; + +netcf_error: + if (driverState->netcf) + { + ncf_close(driverState->netcf); + } + virMutexDestroy (&driverState->lock); +mutex_error: + VIR_FREE(driverState); +alloc_error: + return -1; +} + +static int interfaceCloseInterface(virConnectPtr conn) +{ + + if (conn->interfacePrivateData != NULL) + { + struct interface_driver *driver + = (struct interface_driver *)conn->interfacePrivateData; + + /* close netcf instance */ + ncf_close(driver->netcf); + /* destroy lock */ + virMutexDestroy(&driver->lock); + /* free driverState */ + VIR_FREE(driver); + } + conn->interfacePrivateData = NULL; + return 0; +} + +static int interfaceNumInterfaces(virConnectPtr conn) +{ + int count; + struct interface_driver *driver = conn->interfacePrivateData; + + interfaceDriverLock(driver); + count = ncf_num_of_interfaces(driver->netcf, NETCF_FLAG_ACTIVE); + interfaceDriverUnlock(driver); + return count; +} + +static int interfaceListInterfaces(virConnectPtr conn, char **const names, int nnames) +{ + struct interface_driver *driver = conn->interfacePrivateData; + int count; + + interfaceDriverLock(driver); + + count = ncf_list_interfaces(driver->netcf, nnames, names, NETCF_FLAG_ACTIVE); + /* netcf cleans up in the case of errors */ + interfaceDriverUnlock(driver); + + return count; + +} + +static virInterfacePtr interfaceLookupByName(virConnectPtr conn, + const char *name) +{ + struct interface_driver *driver = conn->interfacePrivateData; + struct netcf_if *iface; + virInterfacePtr ret = NULL; + + interfaceDriverLock(driver); + iface = ncf_lookup_by_name(driver->netcf, name); + + if (!iface) { + interfaceReportError(conn, NULL, NULL, VIR_ERR_NO_INTERFACE, + "%s", _("no interface with matching name")); + goto cleanup; + } + + ret = virGetInterface(conn, ncf_if_name(iface), ncf_if_mac_string(iface)); + +cleanup: + ncf_if_free(iface); + interfaceDriverUnlock(driver); + return ret; +} + +static virInterfacePtr interfaceLookupByMACString(virConnectPtr conn, + const char *macstr) +{ + struct interface_driver *driver = conn->interfacePrivateData; + struct netcf_if *iface; + virInterfacePtr ret = NULL; + + interfaceDriverLock(driver); + iface = ncf_lookup_by_mac_string(driver->netcf, macstr); + + if (!iface) { + interfaceReportError(conn, NULL, NULL, VIR_ERR_NO_INTERFACE, + "%s", _("no interface with matching MAC address")); + goto cleanup; + } + + ret = virGetInterface(conn, ncf_if_name(iface), ncf_if_mac_string(iface)); + +cleanup: + ncf_if_free(iface); + interfaceDriverUnlock(driver); + return ret; +} + +static char *interfaceGetXMLDesc(virInterfacePtr ifinfo, + unsigned int flags ATTRIBUTE_UNUSED) +{ + struct interface_driver *driver = ifinfo->conn->interfacePrivateData; + struct netcf_if *iface = NULL; + char *ret = NULL; + + interfaceDriverLock(driver); + iface = ncf_lookup_by_name(driver->netcf, ifinfo->name); + + if (!iface) { + /* May want to check by name here, but probably pointless */ + interfaceReportError(ifinfo->conn, NULL, ifinfo, VIR_ERR_INVALID_INTERFACE, + "%s", _("no interface with matching MAC Address")); + goto cleanup; + } + + ret = ncf_if_xml_desc(iface); + if (!ret) { + interfaceReportError(ifinfo->conn, NULL, ifinfo, VIR_ERR_INVALID_INTERFACE, + "%s", _("could not get interface XML description")); + goto cleanup; + } + +cleanup: + ncf_if_free(iface); + interfaceDriverUnlock(driver); + return ret; +} + +static virInterfacePtr interfaceDefineXML(virConnectPtr conn, + const char *xml, + unsigned int flags ATTRIBUTE_UNUSED) +{ + struct interface_driver *driver = conn->interfacePrivateData; + struct netcf_if *iface = NULL; + virInterfacePtr ret = NULL; + + interfaceDriverLock(driver); + + /* + * This is where we will want to validate the XML, and possibly + * transform it before sending it on. + */ + + iface = ncf_define(driver->netcf, xml); + if (!iface) { + interfaceReportError(conn, NULL, NULL, VIR_ERR_INVALID_INTERFACE, + "%s", _("failed to define new interface from XML")); + goto cleanup; + } + + ret = virGetInterface(conn, ncf_if_name(iface), ncf_if_mac_string(iface)); + +cleanup: + ncf_if_free(iface); + interfaceDriverUnlock(driver); + return ret; +} + +static int interfaceUndefine(virInterfacePtr ifinfo) { + struct interface_driver *driver = ifinfo->conn->interfacePrivateData; + struct netcf_if *iface = NULL; + int ret = -1; + + interfaceDriverLock(driver); + + iface = ncf_lookup_by_name(driver->netcf, ifinfo->name); + if (!iface) { + interfaceReportError(ifinfo->conn, NULL, ifinfo, VIR_ERR_INVALID_INTERFACE, + "%s", _("no interface with matching MAC address")); + goto cleanup; + } + + ret = ncf_if_undefine(iface); + if (ret < 0) { + interfaceReportError(ifinfo->conn, NULL, ifinfo, VIR_ERR_INVALID_INTERFACE, + "%s", _("failed to undefine interface")); + goto cleanup; + } + +cleanup: + ncf_if_free(iface); + interfaceDriverUnlock(driver); + return ret; +} + +static int interfaceCreate(virInterfacePtr ifinfo, + unsigned int flags ATTRIBUTE_UNUSED) +{ + struct interface_driver *driver = ifinfo->conn->interfacePrivateData; + struct netcf_if *iface = NULL; + int ret = -1; + + interfaceDriverLock(driver); + + iface = ncf_lookup_by_name(driver->netcf, ifinfo->name); + if (!iface) { + interfaceReportError(ifinfo->conn, NULL, ifincof, VIR_ERR_INVALID_INTERFACE, + "%s", _("no interface with matching MAC address")); + goto cleanup; + } + + ret = ncf_if_up(iface); + if (ret < 0) { + interfaceReportError(ifinfo->conn, NULL, ifinfo, VIR_ERR_INVALID_INTERFACE, + "%s", _("failed to create (start) interface")); + goto cleanup; + } + +cleanup: + ncf_if_free(iface); + interfaceDriverUnlock(driver); + return ret; +} + +static int interfaceDestroy(virInterfacePtr ifinfo, + unsigned int flags ATTRIBUTE_UNUSED) +{ + struct interface_driver *driver = ifinfo->conn->interfacePrivateData; + struct netcf_if *iface = NULL; + int ret = -1; + + interfaceDriverLock(driver); + + iface = ncf_lookup_by_name(driver->netcf, ifinfo->name); + if (!iface) { + interfaceReportError(ifinfo->conn, NULL, ifinfo, VIR_ERR_INVALID_INTERFACE, + "%s", _("no interface with matching MAC address")); + goto cleanup; + } + + ret = ncf_if_down(iface); + if (ret < 0) { + interfaceReportError(ifinfo->conn, NULL, ifinfo, VIR_ERR_INVALID_INTERFACE, + "%s", _("failed to destroy (stop) interface")); + goto cleanup; + } + +cleanup: + ncf_if_free(iface); + interfaceDriverUnlock(driver); + return ret; +} + +static virInterfaceDriver interfaceDriver = { + "Interface", + interfaceOpenInterface, /* open */ + interfaceCloseInterface, /* close */ + interfaceNumInterfaces, /* numOfInterfaces */ + interfaceListInterfaces, /* listInterfaces */ + interfaceLookupByName, /* interfaceLookupByName */ + interfaceLookupByMACString, /* interfaceLookupByMACSTring */ + interfaceGetXMLDesc, /* interfaceGetXMLDesc */ + interfaceDefineXML, /* interfaceDefineXML */ + interfaceUndefine, /* interfaceUndefine */ + interfaceCreate, /* interfaceCreate */ + interfaceDestroy, /* interfaceDestroy */ +}; + +int interfaceRegister(void) { + virRegisterInterfaceDriver(&interfaceDriver); + return 0; +} diff --git a/src/interface_driver.h b/src/interface_driver.h new file mode 100644 index 0000000..be00e0b --- /dev/null +++ b/src/interface_driver.h @@ -0,0 +1,34 @@ +/* + * interface_driver.h: core driver methods for managing physical host interfaces + * + * Copyright (C) 2006, 2007 Red Hat, Inc. + * Copyright (C) 2006 Daniel P. Berrange + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Author: Daniel P. Berrange <berrange(a)redhat.com> + */ + + +#ifndef __VIR_INTERFACE__DRIVER_H +#define __VIR_INTERFACE__DRIVER_H + +#include <config.h> + +#include "internal.h" + +int interfaceRegister(void); + +#endif /* __VIR_INTERFACE__DRIVER_H */ -- 1.6.0.6

2 1

[libvirt] VMware ESX driver status update
by Matthias Bolte 26 Jun '09

26 Jun '09

Hi, I'm still working on the VMX config to domain XML mapping for dump/create XML, but it's not complete yet. I got distracted by other urgent, Uni related work that had to be done first. So I won't be able to complete the dump/create XML implementation until tomorrow (freeze for the 0.6.5 release) that was requested in order to get the driver merged into the official repository. But well, with the short release cycle of libvirt not much is lost :) Some details on the VMX config to domain XML mapping: I claimed before that I would need to read most information needed for dump XML from the VMX config file. That's not true. Possibly all information can be retrieved via VI API, but the information is scattered in various places in the VI API object model. I'm currently heading for reading this information from the VMX config file, because all needed information is concentrated in this file. Also, if one changes properties of a virtual machine via VI API and this properties is reflected in the VMX config, the ESX host updates the VMX config, so the information is kept in sync between the object model and the config file. I guess, that the VMX config files contains enough information to fill all or at least most fields of a virDomainDef. So the first goal is to fill a virDomainDef for dump XML. The next goal would be a basic create XML. I claimed before that I would need to write an VMX config file to the ESX host in order to create a new virtual machine, but as I dig trough the VI API in more detail, it seems that this claim maybe false. I'll just have to test this, but haven't had time to do it yet. One problem here is the essential guestOS field of the VMX config, see http://sanbarrow.com/vmx/vmx-guestos.html . For a first try I would set it to 'other' by default, because there is currently no field available in the domain XML to map this information to. But to allow the user to set this filed, I would want to extend to domain XML definition in order to reuse existing code. So how would I do this? Currently I'm just using the virDomainDef struct and the related parse and format functions. One option would be to add a guest field to the virDomainOSDef struct and extend the parse and format functions to handle it. The parse and format functions take flags already, so a flag could be added to indicate if the guest field should be handled or not (just like the VMX extension for the virConfParser). Regards, Matthias

4 5

[libvirt] [PATCH 0/3] Control Linux capabilities in libvirt
by Daniel P. Berrange 26 Jun '09

26 Jun '09

The libvirtd system instance runs privileged (as root), and likewise so do VMs run from it. This is undesirable for obvious security reasons, even if you do have SELinux available. Linux has a concept of capabilities which are actually what gives 'root' its power. If you take away all capabilities from a process running root, you significantly limit its power. eg if you take away CAP_DAC_OVERRIDE, root can now only read files explicitly owned by root, not any other users. There are about 30-something capabilities at this time - see 'man 7 capabilities' for a full list and examples of what each allows you todo Historically it has needed alot of code to manage/modify capabilities, but there is a new library for managing Linux capabilities recently released, which makes their use significantly easier. This is known as libcap-ng http://people.redhat.com/sgrubb/libcap-ng/ I anticipate a several step process 1. Drop all capabilities from spawned processes which don't need it 2. Reduce capabilities from the libvirtd daemon 3. Run VM processes as non-root, unprivileged users 4. Run the libvirtd daemon as as a non-root, semi-privileged user This series of patches lays the very *basic* groundwork for controlling Linux capabilities, doing options 1 and 2. Option '3' is what I'd really like to get working, but it is more invasive, because it will need changes to the QEMU driver so it will chown disks to the 'kvm' user before spawning kvm (likewise for PCI devs, USB devs, etc). Step 1 is a mild security benefit, adding step 3 is a significant step forward. Step 2 doesn't have much tangible benefit, since our functionality requires that we need CAP_DAC_OVERRIDE, CAP_SYS_ADMIN and CAP_NET_ADMIN. Once you give a process those, its pretty much game over for security benefits of removing other capabilities. The 4th option *could* help security, but only if there are some changes to the way distros deploy apps. In particular they'd need to make sure binaries are installed with correct file capabilities, before we could use it, otherwise we wouldnt' be able to get child processes we spawn to keep capabilities they need. In addition, with step 4, we'd be asking admins to trade off functionality against security. eg, we could only drop CAP_DAC_OVERRIDE, if you turned off certain features. It is certainly desirable for users who want to tightly lock down their system, but not for a general purpose install. 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 :|

4 13

[libvirt] [PATCH 5/5] create storage volumes
by Henrik Persson 25 Jun '09

25 Jun '09

This patch does the last magic to be able to create storage volumes when using a disk as backend

3 3

[libvirt] [PATCH] Fix reading storage pool defs on driver startup
by Cole Robinson 25 Jun '09

25 Jun '09

The recent storage refactoring I committed broke reading pool definitions from disk on driver startup: the result was that none of your pools would be defined. Frankly not sure how I managed to miss the brokenness, sorry about that! Also refactor the startup code to match the existing network and domain conventions. Thanks, Cole

2 1