March 2009 - Devel - libvirt List Archives

[libvirt] KVM update
by Zvi Dubitzky 30 Mar '09

30 Mar '09

I am trying to update the kvm driver and userspace using existing kernel 1) using the tar ball from kvm at sourceforge I can do it , but only kvm -84 is there 2) if i try the latest version from the git tree I am able to do the first 2 steps (make --kerneldir=<exisitng kernel headers> , make sync ) . but the final make fails during compile of libkvm.c. results in the file below (dated 3/29/09) Am I missing something Should I recompile the whole kerenl as a 3) Was the live migrate fixed and tested in the latest kernel version thanks Zvi Dubitzky Virtualization and System Architecture Email:dubi@il.ibm.com IBM Haifa Research Laboratory Phone: +972-4-8296182 Haifa, 31905, ISRAEL

3 2

[libvirt] "invalid connection pointer in virConnectClose" from Python bindings
by Charles Duffy 30 Mar '09

30 Mar '09

Howdy. I'm calling libvirt from a program which occasionally has cause to fork() without an immediate exec(). For the sake of simplicity, I presently call close() on all my virConnect objects [which I then delete] before forking and create new ones later. (I'm not forcing an explicit pre-fork garbage collection at present -- hopefully the close() should make one unnecessary). However, libvirt complains (and very occasionally segfaults) as I try to close the connections: libvir: error : invalid connection pointer in virConnectClose libvir: error : invalid connection pointer in virConnectClose Are there any practices I should be following to avoid this? Alternatively, if it is likely to be related to an issue in the Python bindings, is there something I could do to diagnose? Thanks!

2 1

[libvirt] ANNOUNCE: Perl binding Sys-Virt release 0.2.0
by Daniel P. Berrange 28 Mar '09

28 Mar '09

I've made a long overdue update to the Perl binding for libvirt, Sys-Virt. http://search.cpan.org/CPAN/authors/id/D/DA/DANBERR/Sys-Virt-0.2.0.tar.gz The major changes since last time - Fix network create API, and UUID lookups - Implement storage pool, storage vol, node device, security model, domain events and event loop APIs - Improve way constants are exposed to Perl layer - Fix horrible memory leak in methods returning a hash - Fix integer overflow in APIs using 64-bit ints (aka 'long long') - Minimum required libvirt C library for building is 0.6.1 There is comprehensive API documentation available online http://search.cpan.org/dist/Sys-Virt/ As well as a number of example scripts http://cpansearch.perl.org/src/DANBERR/Sys-Virt-0.2.0/examples/ The master source repository for the Perl binding is now at http://libvirt.org/hg/libvirt-perl/ Thanks to everyone who reported bugs, tested the snapshots and sent patches for this release 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 2/2] VirtualBox support to libvirt
by Pritesh Kothari 27 Mar '09

27 Mar '09

Hi All, I have attached a patch which when applied on the HEAD as of today would allow virtualbox support in libvirt. The patch works very well with the VirtualBox OSE version and the 2.2 Beta release. [PATCH 1/2] contains diff of files already in libvirt. [PATCH 2/2] contains new files needed for VirtualBox support. Regards, -pritesh

3 5

[libvirt] libvirt Centos4
by Gian Mario 27 Mar '09

27 Mar '09

Hi to all, i need to compile libvirt-3.3.0 to Centos 4.7. Searching in mailing list i have seen a tread that help me to solve few problems. Actually i face with a error concern glibc, when i run make in libvirt directory, I obtain: /usr/lib64/libgcrypt.so.11: undefined reference to `__read_chk(a)GLIBC_2.4' /usr/lib/gcc/x86_64-redhat-linux/3.4.6/../../../../lib64/libgnutls.so: undefined reference to `__stack_chk_fail(a)GLIBC_2.4' collect2: ld returned 1 exit status make[2]: *** [virsh] Error 1 make[2]: Leaving directory `/root/XEN/libvirt-0.3.3/src' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/root/XEN/libvirt-0.3.3' make: *** [all] Error 2 Any one have some experience on compile libvirt on centos4? there is any possibilities to build it without upgrade glibc? tnks to any help. Ciao e statevi bene. GMario

3 2

[libvirt] PATCH: Support memory ballooning for QEMU
by Daniel P. Berrange 27 Mar '09

27 Mar '09

This patch implements full support for memory ballooning in the QEMU driver. - Fix qemudBuildCommandLine() to set the initial boot time VM memory allocation based off the 'maxmem' config field. - Add call to 'qemudDomainSetMemoryBalloon' immediately at startup to make the guest balloon to initial requested allocation. - In the qemudDomainGetInfo() and qemudDomainDumpXML calls, query the monitor to find current balloon allocation and update config - Implement qemudDomainSetMemory() for running guests using the monitor balloon command In all of these scenarios, if the QEMU being used is too old to support the memory balloon device, the user will get a suitable error or it'll report the statically defined memory allocation from boot time. Daniel diff -r 4bfef84f4d6f src/qemu_conf.c --- a/src/qemu_conf.c Tue Mar 10 12:09:41 2009 +0000 +++ b/src/qemu_conf.c Tue Mar 10 16:16:59 2009 +0000 @@ -870,7 +870,11 @@ int qemudBuildCommandLine(virConnectPtr } \ } while (0) - snprintf(memory, sizeof(memory), "%lu", vm->def->memory/1024); + /* Set '-m MB' based on maxmem, because the lower 'memory' limit + * is set post-startup using the balloon driver. If balloon driver + * is not supported, then they're out of luck anyway + */ + snprintf(memory, sizeof(memory), "%lu", vm->def->maxmem/1024); snprintf(vcpus, sizeof(vcpus), "%lu", vm->def->vcpus); snprintf(domid, sizeof(domid), "%d", vm->def->id); pidfile = virFilePid(driver->stateDir, vm->def->name); diff -r 4bfef84f4d6f src/qemu_driver.c --- a/src/qemu_driver.c Tue Mar 10 12:09:41 2009 +0000 +++ b/src/qemu_driver.c Tue Mar 10 16:16:59 2009 +0000 @@ -122,6 +122,9 @@ static int qemudMonitorCommandExtra(cons const char *extra, const char *extraPrompt, char **reply); +static int qemudDomainSetMemoryBalloon(virConnectPtr conn, + virDomainObjPtr vm, + unsigned long newmem); static struct qemud_driver *qemu_driver = NULL; @@ -1480,6 +1483,7 @@ static int qemudStartVMDaemon(virConnect (qemudDetectVcpuPIDs(conn, vm) < 0) || (qemudInitCpus(conn, vm, migrateFrom) < 0) || (qemudInitPasswords(conn, driver, vm) < 0) || + (qemudDomainSetMemoryBalloon(conn, vm, vm->def->memory) < 0) || (qemudSaveDomainStatus(conn, qemu_driver, vm) < 0)) { qemudShutdownVMDaemon(conn, driver, vm); return -1; @@ -2409,6 +2413,98 @@ cleanup: return ret; } + +#define BALLOON_PREFIX "balloon: actual=" + +/* + * Returns: 0 if balloon not supported, +1 if balloon query worked + * or -1 on failure + */ +static int qemudDomainGetMemoryBalloon(virConnectPtr conn, + virDomainObjPtr vm, + unsigned long *currmem) { + char *reply = NULL; + int ret = -1; + char *offset; + + if (qemudMonitorCommand(vm, "info balloon", &reply) < 0) { + qemudReportError(conn, dom, NULL, VIR_ERR_OPERATION_FAILED, + "%s", _("could not query memory balloon allocation")); + goto cleanup; + } + + /* If the command failed qemu prints: + * device not found, device is locked ... + * No message is printed on success it seems */ + DEBUG ("balloon reply: '%s'", reply); + if ((offset = strstr(reply, BALLOON_PREFIX)) != NULL) { + unsigned int memMB; + char *end; + offset += strlen(BALLOON_PREFIX); + if (virStrToLong_ui(offset, &end, 10, &memMB) < 0) { + qemudReportError(conn, dom, NULL, VIR_ERR_OPERATION_FAILED, + "%s", _("could not parse memory balloon allocation")); + goto cleanup; + } + *currmem = memMB * 1024; + ret = 1; + } else { + /* We don't raise an error here, since its to be expected that + * many QEMU's don't support ballooning + */ + ret = 0; + } + +cleanup: + VIR_FREE(reply); + return ret; +} + +/* + * Returns: 0 if balloon not supported, +1 if balloon query worked + * or -1 on failure + */ +static int qemudDomainSetMemoryBalloon(virConnectPtr conn, + virDomainObjPtr vm, + unsigned long newmem) { + char *cmd; + char *reply = NULL; + int ret = -1; + + /* + * 'newmem' is in KB, QEMU monitor works in MB, and we all wish + * we just worked in bytes with unsigned long long everywhere. + */ + if (virAsprintf(&cmd, "balloon %lu", (newmem / 1024)) < 0) { + virReportOOMError(conn); + goto cleanup; + } + + if (qemudMonitorCommand(vm, cmd, &reply) < 0) { + qemudReportError(conn, dom, NULL, VIR_ERR_OPERATION_FAILED, + "%s", _("could not balloon memory allocation")); + VIR_FREE(cmd); + goto cleanup; + } + VIR_FREE(cmd); + + /* If the command failed qemu prints: + * device not found, device is locked ... + * No message is printed on success it seems */ + DEBUG ("balloon reply: %s", reply); + if (strstr(reply, "\nunknown command:")) { + /* Don't set error - it is expected memory balloon fails on many qemu */ + ret = 0; + } else { + ret = 1; + } + +cleanup: + VIR_FREE(reply); + return ret; +} + + static int qemudDomainSetMemory(virDomainPtr dom, unsigned long newmem) { struct qemud_driver *driver = dom->conn->privateData; virDomainObjPtr vm; @@ -2426,20 +2522,21 @@ static int qemudDomainSetMemory(virDomai goto cleanup; } + if (newmem > vm->def->maxmem) { + qemudReportError(dom->conn, dom, NULL, VIR_ERR_INVALID_ARG, + "%s", _("cannot set memory higher than max memory")); + goto cleanup; + } + if (virDomainIsActive(vm)) { - qemudReportError(dom->conn, dom, NULL, VIR_ERR_NO_SUPPORT, - "%s", _("cannot set memory of an active domain")); - goto cleanup; - } - - if (newmem > vm->def->maxmem) { - qemudReportError(dom->conn, dom, NULL, VIR_ERR_INVALID_ARG, - "%s", _("cannot set memory higher than max memory")); - goto cleanup; - } - - vm->def->memory = newmem; - ret = 0; + ret = qemudDomainSetMemoryBalloon(dom->conn, vm, newmem); + if (ret == 0) + qemudReportError(dom->conn, dom, NULL, VIR_ERR_NO_SUPPORT, + "%s", _("cannot set memory of an active domain")); + } else { + vm->def->memory = newmem; + ret = 0; + } cleanup: if (vm) @@ -2452,6 +2549,8 @@ static int qemudDomainGetInfo(virDomainP struct qemud_driver *driver = dom->conn->privateData; virDomainObjPtr vm; int ret = -1; + int err; + unsigned long balloon; qemuDriverLock(driver); vm = virDomainFindByUUID(&driver->domains, dom->uuid); @@ -2473,8 +2572,15 @@ static int qemudDomainGetInfo(virDomainP } } + err = qemudDomainGetMemoryBalloon(dom->conn, vm, &balloon); + if (err < 0) + goto cleanup; + info->maxMem = vm->def->maxmem; - info->memory = vm->def->memory; + if (err == 0) /* Balloon not supported */ + info->memory = vm->def->memory; + else + info->memory = balloon; info->nrVirtCpu = vm->def->vcpus; ret = 0; @@ -3169,16 +3275,25 @@ static char *qemudDomainDumpXML(virDomai struct qemud_driver *driver = dom->conn->privateData; virDomainObjPtr vm; char *ret = NULL; - - qemuDriverLock(driver); - vm = virDomainFindByUUID(&driver->domains, dom->uuid); - qemuDriverUnlock(driver); - - if (!vm) { - qemudReportError(dom->conn, dom, NULL, VIR_ERR_INVALID_DOMAIN, - "%s", _("no domain with matching uuid")); - goto cleanup; - } + unsigned long balloon; + int err; + + qemuDriverLock(driver); + vm = virDomainFindByUUID(&driver->domains, dom->uuid); + qemuDriverUnlock(driver); + + if (!vm) { + qemudReportError(dom->conn, dom, NULL, VIR_ERR_INVALID_DOMAIN, + "%s", _("no domain with matching uuid")); + goto cleanup; + } + + /* Refresh current memory based on balloon info */ + err = qemudDomainGetMemoryBalloon(dom->conn, vm, &balloon); + if (err < 0) + goto cleanup; + if (err > 0) + vm->def->memory = balloon; ret = virDomainDefFormat(dom->conn, (flags & VIR_DOMAIN_XML_INACTIVE) && vm->newDef ? -- |: 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 :|

3 4

[libvirt] Libvirt - Problems in different Versions
by Artur Baruchi 27 Mar '09

27 Mar '09

Hi Guys. Im writing a program that connects to Xen and gets some information (scheduller and memory informations). When executing the program using the libvirt 0.3.3 its working fine, BUT when running using libvirt 0.4.4 it doesnt work. I dont get any error message when compiling or executing, but the informations doesnt come. -> Using libvirt 0.3.3 # gcc -Wall -lxenstore `pkg-config --cflags --libs libvirt` test.c -o test # ./test DOMAINS IDENTIFIED 2 DomID: 0 Max Mem: 4294967292 Sched: 256 DomID: 3 Max Mem: 262144 Sched: 256 # xm list Name ID Mem VCPUs State Time(s) Domain-0 0 616 2 r----- 4751.9 vm01 3 256 2 -b---- 379.7 -> Using libvirt 0.4.4 # gcc -Wall -lxenstore `pkg-config --cflags --libs libvirt` test.c -o test # ./test DOMAINS IDENTIFIED 0 # xm list Name ID Mem VCPUs State Time(s) Domain-0 0 481 2 r----- 89.8 linuxprv01 1 512 1 -b---- 10.8 I executed the same program in other server, with 0.4.4 version, and got the same problem. Is there any difference between those versions? Thanks in advance. Att. Artur Baruchi

2 1

[libvirt] Using libvirt from PHP
by Radek Hladik 26 Mar '09

26 Mar '09

Hi, I am developing wimple web application in PHP to monitor and control VMs using libvirt. I hope it could become simple library for doing basic things with VMs. And I would like to ask for your opinion about different possibilities of calling libvirt. Option 1) I am calling virsh via exec command now (one virsh for every libvirt command). This can be used for local libvirt but it is very inefficient. The overhead for remote libvirt would be unbearable. I could run one virsh per page and use it "interactively" so the overhead gets lower. Option 2) The other option is to create Zend extension which should be able to call libvirt directly from C. However I've never done anything like that so I do not know, whether it is reasonable or not. Option 3) I could also create some "wrapper daemon" in C or other language with libvirt binding. But I think that this is an ugly way :-) Option 4) I could "talk" directly to the libvirt socket. But I am not sure how the communication goes there and whether it is stable or changes with every version. And I consider the socket to be internal thing of libvirt... Is there any other option I missed? And which one would you suggest? Radek

5 5

[libvirt] Monitoring VMs via java application
by Daniel Halle 26 Mar '09

26 Mar '09

Hello all, I want to monitor my VMs with a Java application. Is there a possibility to register an observer or the like which is triggered e.g. after the state of a VM has changed? Or is the only way to use a timer which polls all VMs every xy minutes? Thanks, Daniel -- Infinis GmbH - Taläckerstr. 30 - D-70437 Stuttgart Geschäftsführer: Daniel Schwager, Stefan Hörz - HRB Stuttgart 20964 Tel: 0711-849910-23 - Fax: 0711-849910-923 - daniel.halle(a)infinis.de

2 1

[libvirt] Restoring from a largish 'virsh save' file
by Matt McCowan 26 Mar '09

26 Mar '09

Running into an issue where, if I/O is hampered by load for example, reading a largish state file (created by 'virsh save') is not allowed to complete. qemudStartVMDaemon in src/qemu_driver.c has a loop that waits 10 seconds for the VM to be brought up. An strace against libvirt when doing a 'virsh restore' against a largish state file shows the VM being sent a kill when it's still happily reading from the file. inotify is used in the xen and uml drivers, so I thought it would be a suitable mechanism to delay the timeout loop if the state file was still being read. Is this the right way to solve this problem? I've inlined some (flimsy) mods for comment Regards Matt McCowan --- libvirt.orig/src/qemu_driver.c 2009-03-20 11:17:57.000000000 +0900 +++ libvirt-0.6.1.1/src/qemu_driver.c 2009-03-22 18:36:06.000000000 +0900 @@ -43,6 +43,7 @@ #include <stdio.h> #include <sys/wait.h> #include <sys/ioctl.h> +#include <sys/inotify.h> #if HAVE_NUMACTL #define NUMA_VERSION1_COMPATIBILITY 1 @@ -1310,6 +1309,11 @@ pid_t child; int pos = -1; char ebuf[1024]; + char rpathname[PATH_MAX]; + int fd = 0; + int wd = 0; + int got = 0; + char buf[1024]; struct gemudHookData hookData; hookData.conn = conn; @@ -1430,6 +1434,19 @@ qemudSecurityHook, &hookData); /* wait for qemu process to to show up */ + + /* Times out if using a restore file and that file + is big. Use inotify */ + if (stdin_fd > 0) { + fd = inotify_init(); + + if (snprintf(rpathname, sizeof(rpathname), "/proc/self/fd/%d", + stdin_fd) >= (int)sizeof(rpathname)) { + return -1; + } + wd = inotify_add_watch(fd, rpathname, IN_ACCESS); + } + if (ret == 0) { int retries = 100; int childstat; @@ -1442,7 +1459,15 @@ if ((ret = virFileReadPid(driver->stateDir, vm->def->name, &vm->pid)) == 0) break; usleep(100*1000); - retries--; + + /* Is state file being accessed? */ + got = read(fd, buf, sizeof(buf)); + if (got < 1) + retries--; + } + if (stdin_fd > 0) { + inotify_rm_watch(fd, wd); + close(fd); } if (ret) { qemudReportError(conn, NULL, NULL, VIR_ERR_INTERNAL_ERROR,

2 3