July 2008 - Devel - libvirt List Archives

[libvirt] auto shutdown for qemu / kvm
by Matthias Pfafferodt 15 Jul '08

15 Jul '08

Hello, there is the possibility to autostart a (qemu/kvm) domain. If libvirtd is stopped this domain is killed and the filesystem is in a unclean state. I miss the possible of an 'autostop' option that will shutdown a (qemu/kvm) domain if libvirtd is stopped. I changed the init script of my system (opensuse 11.0) to implement the 'autostop' option. Is there a simpler solution? Do you have tips how to improve the script? Matthias ## Path: System/Virtualization ## Description: libvirtd stop configuration ## Type: yesno ## Default: yes # # On 'libvirtd stop' all running (kvm) domains are killed. If # LIBVIRTD_KVM_SHUTDOWN is set to 'yes' all running vm's are given time # for a gracefully shutdown. # LIBVIRTD_KVM_SHUTDOWN="yes" ## Path: System/Virtualization ## Description: libvirtd timeout ## Type: integer ## Default: 300 # # On 'libvirtd stop' the vm's are ask to shutdown. This optine gives the # maximum number of seconds the script will wait until libvirt is stopped. # LIBVIRTD_KVM_SHUTDOWN_MAXWAIT=300

2 2

[libvirt] static address assignment via dnsmasq
by Kurt Sussman 15 Jul '08

15 Jul '08

I want my VMs to be assigned static addresses by hostname. I would prefer to set up a generic way to configure dnsmasq, since I don't think this is the last thing I'll want to change about address assignment. I saw this patch from "Mads Chr. Olesen" https://www.redhat.com/archives/libvir-list/2008-April/msg00327.html But it seems like a lot of work to give access to only one dnsmasq option. I'm working on a patch that enables a dnsmasq.d directory where you can (optionally) put individual configuration files. I have not fully tested it yet, so please don't consider this an official submission. I also don't feel that I have spent enough time with the code to understand the 'philosphy' if you will, the guiding attitudes. Is the all-XML approach better according the libvirt developer community standards? If so, why? Thanks. Untested patch follows. Please let me know if I'm on the right path and if not, why. --Kurt diff -u /root/libvirt-0.4.4-virgin/src/qemu_driver.c ./qemu_driver.c --- /root/libvirt-0.4.4-virgin/src/qemu_driver.c 2008-06-12 07:52:53.000000000 -0700 +++ ./qemu_driver.c 2008-07-13 11:31:26.000000000 -0700 @@ -1048,6 +1048,7 @@ 1 + /* --bind-interfaces */ 2 + /* --pid-file "" */ 2 + /* --conf-file "" */ + 2 + /* --conf-dir DNSMASQCONF_DIR */ /*2 + *//* --interface virbr0 */ 2 + /* --except-interface lo */ 2 + /* --listen-address 10.0.0.1 */ @@ -1081,6 +1082,9 @@ APPEND_ARG(*argv, i++, "--conf-file"); APPEND_ARG(*argv, i++, ""); + APPEND_ARG(*argv, i++, "--conf-dir"); + APPEND_ARG(*argv, i++, DNSMASQCONF_DIR); + /* * XXX does not actually work, due to some kind of * race condition setting up ipv6 addresses on the diff -u /root/libvirt-0.4.4-virgin/src/qemu_driver.h ./qemu_driver.h --- /root/libvirt-0.4.4-virgin/src/qemu_driver.h 2008-04-23 07:19:02.000000000 -0700 +++ ./qemu_driver.h 2008-07-13 11:26:47.000000000 -0700 @@ -33,6 +33,8 @@ int qemudRegister(void); +#define DNSMASQCONF_DIR SYSCONF_DIR "/dnsmasq.d" + #endif /* WITH_QEMU */ #endif /* QEMUD_DRIVER_H */ -- ---------------------------------------------------------------------- Merlot Research Group, Inc kls[at]merlot.com "Quality is free, but only to those who are willing to pay heavily for it." (Lister, DeMarco: "Peopleware")

2 1

[libvirt] PATCH: Refactor LXC to remove some state from lxc_vm_t
by Daniel P. Berrange 14 Jul '08

14 Jul '08

I've been working on getting LXC ready to use the common XML routines. One of the first tasks is finding a way to remove some of the state stored in the lxc_vm_t struct, so it can be switched over to virDomainObj struct instead. This patch does some simple re-factoring of the way the TTYs and control socket are handled to prepare for the switchover. lxc_conf.c | 1 lxc_conf.h | 10 -- lxc_container.c | 61 +++++++--------- lxc_container.h | 14 +++ lxc_driver.c | 211 ++++++++++++++++---------------------------------------- 5 files changed, 103 insertions(+), 194 deletions(-) Daniel diff -r 6273f6456bd2 src/lxc_conf.c --- a/src/lxc_conf.c Sun Jul 13 10:33:25 2008 +0100 +++ b/src/lxc_conf.c Sun Jul 13 11:45:04 2008 +0100 @@ -1041,7 +1041,6 @@ void lxcFreeVM(lxc_vm_t *vm) { lxcFreeVMDef(vm->def); - VIR_FREE(vm->containerTty); VIR_FREE(vm); } diff -r 6273f6456bd2 src/lxc_conf.h --- a/src/lxc_conf.h Sun Jul 13 10:33:25 2008 +0100 +++ b/src/lxc_conf.h Sun Jul 13 11:45:04 2008 +0100 @@ -35,12 +35,6 @@ #define LXC_MAX_XML_LENGTH 16384 #define LXC_MAX_ERROR_LEN 1024 #define LXC_DOMAIN_TYPE "lxc" -#define LXC_PARENT_SOCKET 0 -#define LXC_CONTAINER_SOCKET 1 - -/* messages between parent and container */ -typedef char lxc_message_t; -#define LXC_CONTINUE_MSG 'c' /* types of networks for containers */ enum lxc_net_type { @@ -100,10 +94,6 @@ char ttyPidFile[PATH_MAX]; int parentTty; - int containerTtyFd; - char *containerTty; - - int sockpair[2]; lxc_vm_def_t *def; diff -r 6273f6456bd2 src/lxc_container.c --- a/src/lxc_container.c Sun Jul 13 10:33:25 2008 +0100 +++ b/src/lxc_container.c Sun Jul 13 11:45:04 2008 +0100 @@ -33,7 +33,6 @@ #include <unistd.h> #include "lxc_container.h" -#include "lxc_conf.h" #include "util.h" #include "memory.h" #include "veth.h" @@ -83,7 +82,7 @@ * * Returns 0 on success or -1 in case of error */ -static int lxcSetContainerStdio(const char *ttyName) +static int lxcSetContainerStdio(const char *ttyPath) { int rc = -1; int ttyfd; @@ -94,10 +93,10 @@ goto error_out; } - ttyfd = open(ttyName, O_RDWR|O_NOCTTY); + ttyfd = open(ttyPath, O_RDWR|O_NOCTTY); if (ttyfd < 0) { lxcError(NULL, NULL, VIR_ERR_INTERNAL_ERROR, - _("open(%s) failed: %s"), ttyName, strerror(errno)); + _("open(%s) failed: %s"), ttyPath, strerror(errno)); goto error_out; } @@ -106,8 +105,6 @@ _("ioctl(TIOCSTTY) failed: %s"), strerror(errno)); goto cleanup; } - - close(0); close(1); close(2); if (dup2(ttyfd, 0) < 0) { lxcError(NULL, NULL, VIR_ERR_INTERNAL_ERROR, @@ -144,12 +141,18 @@ * * Returns 0 on success or -1 in case of error */ -static int lxcExecWithTty(lxc_vm_t *vm) +static int lxcExecWithTty(lxc_vm_def_t *vmDef, char *ttyPath) { int rc = -1; - lxc_vm_def_t *vmDef = vm->def; + int open_max, i; - if(lxcSetContainerStdio(vm->containerTty) < 0) { + /* Just in case someone forget to set FD_CLOEXEC, explicitly + * close all FDs before executing the container */ + open_max = sysconf (_SC_OPEN_MAX); + for (i = 0; i < open_max; i++) + close(i); + + if(lxcSetContainerStdio(ttyPath) < 0) { goto exit_with_error; } @@ -161,7 +164,7 @@ /** * lxcWaitForContinue: - * @vm: Pointer to vm structure + * @monitor: monitor FD from parent * * This function will wait for the container continue message from the * parent process. It will send this message on the socket pair stored in @@ -169,31 +172,23 @@ * * Returns 0 on success or -1 in case of error */ -static int lxcWaitForContinue(lxc_vm_t *vm) +static int lxcWaitForContinue(int monitor) { - int rc = -1; lxc_message_t msg; int readLen; - readLen = saferead(vm->sockpair[LXC_CONTAINER_SOCKET], &msg, sizeof(msg)); - if (readLen != sizeof(msg)) { + readLen = saferead(monitor, &msg, sizeof(msg)); + if (readLen != sizeof(msg) || + msg != LXC_CONTINUE_MSG) { lxcError(NULL, NULL, VIR_ERR_INTERNAL_ERROR, _("Failed to read the container continue message: %s"), strerror(errno)); - goto error_out; + return -1; } DEBUG0("Received container continue message"); - close(vm->sockpair[LXC_PARENT_SOCKET]); - vm->sockpair[LXC_PARENT_SOCKET] = -1; - close(vm->sockpair[LXC_CONTAINER_SOCKET]); - vm->sockpair[LXC_CONTAINER_SOCKET] = -1; - - rc = 0; - -error_out: - return rc; + return 0; } /** @@ -204,12 +199,12 @@ * * Returns 0 on success or nonzero in case of error */ -static int lxcEnableInterfaces(const lxc_vm_t *vm) +static int lxcEnableInterfaces(const lxc_vm_def_t *def) { int rc = 0; const lxc_net_def_t *net; - for (net = vm->def->nets; net; net = net->next) { + for (net = def->nets; net; net = net->next) { DEBUG("Enabling %s", net->containerVeth); rc = vethInterfaceUpOrDown(net->containerVeth, 1); if (0 != rc) { @@ -218,7 +213,7 @@ } /* enable lo device only if there were other net devices */ - if (vm->def->nets) + if (def->nets) rc = vethInterfaceUpOrDown("lo", 1); error_out: @@ -237,11 +232,11 @@ * * Returns 0 on success or -1 in case of error */ -int lxcChild( void *argv ) +int lxcChild( void *data ) { int rc = -1; - lxc_vm_t *vm = (lxc_vm_t *)argv; - lxc_vm_def_t *vmDef = vm->def; + lxc_child_argv_t *argv = data; + lxc_vm_def_t *vmDef = argv->config; lxc_mount_t *curMount; int i; @@ -278,16 +273,16 @@ } /* Wait for interface devices to show up */ - if (0 != (rc = lxcWaitForContinue(vm))) { + if (0 != (rc = lxcWaitForContinue(argv->monitor))) { goto cleanup; } /* enable interfaces */ - if (0 != (rc = lxcEnableInterfaces(vm))) { + if (0 != (rc = lxcEnableInterfaces(vmDef))) { goto cleanup; } - rc = lxcExecWithTty(vm); + rc = lxcExecWithTty(vmDef, argv->ttyPath); /* this function will only return if an error occured */ cleanup: diff -r 6273f6456bd2 src/lxc_container.h --- a/src/lxc_container.h Sun Jul 13 10:33:25 2008 +0100 +++ b/src/lxc_container.h Sun Jul 13 11:45:04 2008 +0100 @@ -24,7 +24,21 @@ #ifndef LXC_CONTAINER_H #define LXC_CONTAINER_H +#include "lxc_conf.h" + #ifdef WITH_LXC + +typedef struct __lxc_child_argv lxc_child_argv_t; +struct __lxc_child_argv { + lxc_vm_def_t *config; + int monitor; + char *ttyPath; +}; + +/* messages between parent and container */ +typedef char lxc_message_t; +#define LXC_CONTINUE_MSG 'c' + /* Function declarations */ int lxcChild( void *argv ); diff -r 6273f6456bd2 src/lxc_driver.c --- a/src/lxc_driver.c Sun Jul 13 10:33:25 2008 +0100 +++ b/src/lxc_driver.c Sun Jul 13 11:45:04 2008 +0100 @@ -561,27 +561,23 @@ /** * lxcSendContainerContinue: - * @vm: pointer to virtual machine structure + * @monitor: FD for communicating with child * * Sends the continue message via the socket pair stored in the vm * structure. * * Returns 0 on success or -1 in case of error */ -static int lxcSendContainerContinue(const lxc_vm_t *vm) +static int lxcSendContainerContinue(virConnectPtr conn, + int monitor) { int rc = -1; lxc_message_t msg = LXC_CONTINUE_MSG; int writeCount = 0; - if (NULL == vm) { - goto error_out; - } - - writeCount = safewrite(vm->sockpair[LXC_PARENT_SOCKET], &msg, - sizeof(msg)); + writeCount = safewrite(monitor, &msg, sizeof(msg)); if (writeCount != sizeof(msg)) { - lxcError(NULL, NULL, VIR_ERR_INTERNAL_ERROR, + lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, _("unable to send container continue message: %s"), strerror(errno)); goto error_out; @@ -605,12 +601,15 @@ */ static int lxcStartContainer(virConnectPtr conn, lxc_driver_t* driver, - lxc_vm_t *vm) + lxc_vm_t *vm, + int monitor, + char *ttyPath) { int rc = -1; int flags; int stacksize = getpagesize() * 4; char *stack, *stacktop; + lxc_child_argv_t args = { vm->def, monitor, ttyPath }; /* allocate a stack for the container */ if (VIR_ALLOC_N(stack, stacksize) < 0) { @@ -625,7 +624,7 @@ if (vm->def->nets != NULL) flags |= CLONE_NEWNET; - vm->def->id = clone(lxcChild, stacktop, flags, (void *)vm); + vm->def->id = clone(lxcChild, stacktop, flags, &args); DEBUG("clone() returned, %d", vm->def->id); @@ -643,117 +642,9 @@ return rc; } -/** - * lxcPutTtyInRawMode: - * @conn: pointer to connection - * @ttyDev: file descriptor for tty - * - * Sets tty attributes via cfmakeraw() - * - * Returns 0 on success or -1 in case of error - */ -static int lxcPutTtyInRawMode(virConnectPtr conn, int ttyDev) -{ - int rc = -1; - - struct termios ttyAttr; - - if (tcgetattr(ttyDev, &ttyAttr) < 0) { - lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, - "tcgetattr() failed: %s", strerror(errno)); - goto cleanup; - } - - cfmakeraw(&ttyAttr); - - if (tcsetattr(ttyDev, TCSADRAIN, &ttyAttr) < 0) { - lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, - "tcsetattr failed: %s", strerror(errno)); - goto cleanup; - } - - rc = 0; - -cleanup: - return rc; -} /** - * lxcSetupTtyTunnel: - * @conn: pointer to connection - * @vmDef: pointer to virtual machine definition structure - * @ttyDev: pointer to int. On success will be set to fd for master - * end of tty - * - * Opens and configures the parent side tty - * - * Returns 0 on success or -1 in case of error - */ -static int lxcSetupTtyTunnel(virConnectPtr conn, - lxc_vm_def_t *vmDef, - int* ttyDev) -{ - int rc = -1; - char *ptsStr; - - if (0 < strlen(vmDef->tty)) { - *ttyDev = posix_openpt(O_RDWR|O_NOCTTY|O_NONBLOCK); - if (*ttyDev < 0) { - lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, - "open() tty failed: %s", strerror(errno)); - goto setup_complete; - } - - rc = grantpt(*ttyDev); - if (rc < 0) { - lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, - "grantpt() failed: %s", strerror(errno)); - goto setup_complete; - } - - rc = unlockpt(*ttyDev); - if (rc < 0) { - lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, - "unlockpt() failed: %s", strerror(errno)); - goto setup_complete; - } - - /* get the name and print it to stdout */ - ptsStr = ptsname(*ttyDev); - if (ptsStr == NULL) { - lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, - "ptsname() failed"); - goto setup_complete; - } - /* This value needs to be stored in the container configuration file */ - VIR_FREE(vmDef->tty); - if (!(vmDef->tty = strdup(ptsStr))) { - lxcError(conn, NULL, VIR_ERR_NO_MEMORY, - _("unable to get storage for vm tty name")); - goto setup_complete; - } - - /* Enter raw mode, so all characters are passed directly to child */ - if (lxcPutTtyInRawMode(conn, *ttyDev) < 0) { - goto setup_complete; - } - - } else { - *ttyDev = -1; - } - - rc = 0; - -setup_complete: - if((0 != rc) && (*ttyDev > 0)) { - close(*ttyDev); - } - - return rc; -} - -/** - * lxcSetupContainerTty: + * lxcOpenTty: * @conn: pointer to connection * @ttymaster: pointer to int. On success, set to fd for master end * @ttyName: On success, will point to string slave end of tty. Caller @@ -763,12 +654,12 @@ * * Returns 0 on success or -1 in case of error */ -static int lxcSetupContainerTty(virConnectPtr conn, - int *ttymaster, - char **ttyName) +static int lxcOpenTty(virConnectPtr conn, + int *ttymaster, + char **ttyName, + int rawmode) { int rc = -1; - char tempTtyName[PATH_MAX]; *ttymaster = posix_openpt(O_RDWR|O_NOCTTY|O_NONBLOCK); if (*ttymaster < 0) { @@ -783,27 +674,43 @@ goto cleanup; } - if (0 != ptsname_r(*ttymaster, tempTtyName, sizeof(tempTtyName))) { - lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, - _("ptsname_r failed: %s"), strerror(errno)); - goto cleanup; + if (rawmode) { + struct termios ttyAttr; + if (tcgetattr(*ttymaster, &ttyAttr) < 0) { + lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, + "tcgetattr() failed: %s", strerror(errno)); + goto cleanup; + } + + cfmakeraw(&ttyAttr); + + if (tcsetattr(*ttymaster, TCSADRAIN, &ttyAttr) < 0) { + lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, + "tcsetattr failed: %s", strerror(errno)); + goto cleanup; + } } - if (VIR_ALLOC_N(*ttyName, strlen(tempTtyName) + 1) < 0) { - lxcError(conn, NULL, VIR_ERR_NO_MEMORY, - _("unable to allocate container name string")); - goto cleanup; + if (ttyName) { + char tempTtyName[PATH_MAX]; + if (0 != ptsname_r(*ttymaster, tempTtyName, sizeof(tempTtyName))) { + lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, + _("ptsname_r failed: %s"), strerror(errno)); + goto cleanup; + } + + if ((*ttyName = strdup(tempTtyName)) == NULL) { + lxcError(conn, NULL, VIR_ERR_NO_MEMORY, NULL); + goto cleanup; + } } - - strcpy(*ttyName, tempTtyName); rc = 0; cleanup: - if (0 != rc) { - if (-1 != *ttymaster) { - close(*ttymaster); - } + if (rc != 0 && + *ttymaster != -1) { + close(*ttymaster); } return rc; @@ -989,15 +896,18 @@ lxc_vm_t * vm) { int rc = -1; - lxc_vm_def_t *vmDef = vm->def; + int sockpair[2]; + int containerTty; + char *containerTtyPath = NULL; /* open parent tty */ - if (lxcSetupTtyTunnel(conn, vmDef, &vm->parentTty) < 0) { + VIR_FREE(vm->def->tty); + if (lxcOpenTty(conn, &vm->parentTty, &vm->def->tty, 1) < 0) { goto cleanup; } /* open container tty */ - if (lxcSetupContainerTty(conn, &(vm->containerTtyFd), &(vm->containerTty)) < 0) { + if (lxcOpenTty(conn, &containerTty, &containerTtyPath, 0) < 0) { goto cleanup; } @@ -1011,7 +921,7 @@ if (vm->pid == 0) { /* child process calls forward routine */ - lxcTtyForward(vm->parentTty, vm->containerTtyFd); + lxcTtyForward(vm->parentTty, containerTty); } if (lxcStoreTtyPid(driver, vm)) { @@ -1019,7 +929,7 @@ } close(vm->parentTty); - close(vm->containerTtyFd); + close(containerTty); if (0 != (rc = lxcSetupInterfaces(conn, vm))) { goto cleanup; @@ -1027,7 +937,7 @@ /* create a socket pair to send continue message to the container once */ /* we've completed the post clone configuration */ - if (0 != socketpair(PF_UNIX, SOCK_STREAM, 0, vm->sockpair)) { + if (0 != socketpair(PF_UNIX, SOCK_STREAM, 0, sockpair)) { lxcError(conn, NULL, VIR_ERR_INTERNAL_ERROR, _("sockpair failed: %s"), strerror(errno)); goto cleanup; @@ -1035,7 +945,9 @@ /* check this rc */ - rc = lxcStartContainer(conn, driver, vm); + rc = lxcStartContainer(conn, driver, vm, + sockpair[1], + containerTtyPath); if (rc != 0) goto cleanup; @@ -1043,7 +955,7 @@ if (rc != 0) goto cleanup; - rc = lxcSendContainerContinue(vm); + rc = lxcSendContainerContinue(conn, sockpair[0]); if (rc != 0) goto cleanup; @@ -1052,10 +964,9 @@ driver->nactivevms++; cleanup: - close(vm->sockpair[LXC_PARENT_SOCKET]); - vm->sockpair[LXC_PARENT_SOCKET] = -1; - close(vm->sockpair[LXC_CONTAINER_SOCKET]); - vm->sockpair[LXC_CONTAINER_SOCKET] = -1; + close(sockpair[0]); + close(sockpair[1]); + VIR_FREE(containerTtyPath); return rc; } -- |: 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] save / restore for KVM
by Matthias Pfafferodt 12 Jul '08

12 Jul '08

Hello, I try to save a KVM domain which has its hard disk on a LVM volume. I can execute the command to save the domain but restore fails # virsh list Id Name State ---------------------------------- 3 kvm-linux03 running 4 kvm-linux01 running # virsh save kvm-linux01 ./kvm-linux01.ram Domain kvm-linux01 saved to ./kvm-linux01.ram # virsh list Id Name State ---------------------------------- 3 kvm-linux03 running # virsh restore ./kvm-linux01.ram libvir: QEMU error : operation failed: failed to start VM error: Failed to restore domain from ./kvm-linux01.ram I tried to save and restore this domain using kvm directly # /usr/bin/qemu-kvm -m 1024 -boot c -drive file=/dev/virtual/kvm_linux01,if=ide,index=0,boot=on -net nic,macaddr=00:16:3e:00:01:02,vlan=0 -net tap,fd=24,script=,vlan=0,ifname=vnet3 (qemu) savevm kvm-linux01.ram Device id0-hd0 does not support VM state snapshots (qemu) Here I get an error message which I also found on the qemu homepage: to save a qemu vm the hard disk must be a file using the qcow format. But in the libvirt documentation there is no hint to this limitation. Is it possible to save the state of a kvm vm using physical hard disks with libvirt? Kind regards Matthias Pfafferodt -- Matthias Pfafferodt - http://www.mapfa.de Matthias.Pfafferodt <at> mapfa.de

3 4

[libvirt] [PATCH] default URI in probe
by Evgeniy Sokolov 11 Jul '08

11 Jul '08

I specified default URI in probe(). To use libvirt with OpenVZ: # ./configure --with-openvz # make # ./src/virsh list Id Name State ---------------------------------- 33 33 running 105 105 running 108 108 running

2 1

[libvirt] [PATCH] using one function to locate OpenVZ config files
by Evgeniy Sokolov 11 Jul '08

11 Jul '08

There were 3 places with the same code. I moved it to separate function. other: use O_RDONLY to read config file

2 1

[libvirt] bug in libvirt.c?
by Evgeniy Sokolov 10 Jul '08

10 Jul '08

for (i = 0; i < virNetworkDriverTabCount; i++) { if ((virDriverTab[i]->probe != NULL) && ((latest = virDriverTab[i]->probe()) != NULL)) { probes++; ..... } } We use array virDriverTab, but cicle is to virNetworkDriverTabCount which count number of virNetworkDriverTab drivers. I think virDriverTabCount should be used. Index: libvirt.c =================================================================== RCS file: /data/cvs/libvirt/src/libvirt.c,v retrieving revision 1.147 diff -u -p -r1.147 libvirt.c --- libvirt.c 26 Jun 2008 09:37:51 -0000 1.147 +++ libvirt.c 10 Jul 2008 14:48:44 -0000 @@ -699,7 +699,7 @@ do_open (const char *name, const char *use = NULL; const char *latest; int probes = 0; - for (i = 0; i < virNetworkDriverTabCount; i++) { + for (i = 0; i < virDriverTabCount; i++) { if ((virDriverTab[i]->probe != NULL) && ((latest = virDriverTab[i]->probe()) != NULL)) { probes++;

3 2

[libvirt] [Patch] Libvirt & Autobuild
by Mohammed Morsi 10 Jul '08

10 Jul '08

Attached is a small fix to get libvirt working w/ autobuild. Autobuild attempts to create and test the RPM, which lists the PolicyKit as one of the files to be installed, and thus if it is not installed, autobuild reports an error and fails. The was to get this file installed is to specify --with-polkit to ./autogen which I added to the autobuild.sh build script. (also the syntax check errors broke autobuild, but these have been fixed according to the latest email on the list). Knock on wood, after this libvirt should work w/ autobuild and will be built nightly, and I can finish getting oVirt working with it. -Mo

2 3

[libvirt] PV loader in capabilities?
by John Levon 10 Jul '08

10 Jul '08

Why is it NULL instead of the path to pygrub? On Solaris, the path is different. regards john

3 4

[libvirt] [PATCH] fix diagnostic when execute openvz commands
by Evgeniy Sokolov 10 Jul '08

10 Jul '08

-- Evgeniy Sokolov

4 7

[libvirt] [PATCH] renaming error -> openvzError
by Evgeniy Sokolov 10 Jul '08

10 Jul '08

function error() was renamed to openvzError() to don't conflict with other functions with the same name. other - openvzError() is added to msg_gen_function - fixed to pass "make syntax-check" Thanks to Jim Meyering for review.

2 1

[libvirt] [PATCH] fix up dom-name validation
by John Levon 10 Jul '08

10 Jul '08

libvirt schema: improve domain name regex This regexp is based upon what xend actually allows, plus the existing &, which I presume is there for a reason. Signed-off-by: John Levon <john.levon> Index: docs/libvirt.rng =================================================================== RCS file: /data/cvs/libvirt/docs/libvirt.rng,v retrieving revision 1.8 diff -u -r1.8 libvirt.rng --- docs/libvirt.rng 8 Jul 2008 12:05:13 -0000 1.8 +++ docs/libvirt.rng 10 Jul 2008 04:04:58 -0000 @@ -840,7 +840,7 @@ </define> <define name='domainName'> <data type='string'> - <param name="pattern">[a-zA-Z0-9_\+\-&]+</param> + <param name="pattern">[A-Za-z0-9_.:/\+\-&-]+</param> </data> </define> <define name='genericName'>

2 1

[libvirt] [PATCH] SetAutostart and GetAutostart in openvz driver
by Evgeniy Sokolov 10 Jul '08

10 Jul '08

Implemented functions domainSetAutostart and domainGetAutostart. other: fixed closing file descriptor during reading uuid from config.

3 3

[libvirt] virDomainMemoryPeek & maximum remote message buffer size
by Richard W.M. Jones 09 Jul '08

09 Jul '08

The kernel images that I want to snoop in virt-mem are around 16 MB in size. In the qemu / KVM case, these images have to travel over the remote connection. Because of limits on the maximum message size, they have to travel currently in 64 KB chunks, and it turns out that this is slow. Apparently the dominating factors are how long it takes to issue the 'memsave' command in the qemu monitor (there is some big constant overhead), and extra network round trips. The current remote message size is intentionally limited to 256KB (fully serialized, including all XDR headers and overhead), so the most we could practically send in a single message at the moment is 128KB if we stick to powers of two, or ~255KB if we don't. The reason we limit it is to avoid denial of service attacks, where a rogue client or server sends excessively large messages and causes the peer to allocate lots of memory [eg. if we didn't have any limit, then you could send a message which was several GB in size and cause problems at the other end, because the message is slurped in before it is fully parsed]. There is a second problem with reading the kernel in small chunks, namely that this allows the virtual machine to make a lot of progress, so we don't get anything near an 'instantaneous' snapshot (getting the kernel in a single chunk doesn't necessarily guarantee this either, but it's better). As an experiment, I tried increasing the maximum message to 32 MB, so that I could send the whole kernel in one go. Unfortunately just increasing the limit doesn't work for two reasons, one prosaic and one very weird: (1) The current code likes to keep message buffers on the stack, and because Linux limits the stack to something artificially small, this fails. Increasing the stack ulimit is a short-term fix for this, while testing. In the long term we could rewrite any code which does this to use heap buffers instead. (2) There is some really odd problem with our use of recv(2) which causes messages > 64 KB to fail. I have no idea what is really happening, but the sequence of events seems to be this: server client write(sock,buf,len) = len-k recv(sock,buf,len) = len-k write(sock,buf+len-k,len-k) = k recv(sock,buf,k) = 0 [NOT k] At this point the client assumes that the server has unexpectedly closed the connection and fails. I have stared at this for a while, but I've got no idea at all what's going on. A test program is attached. You'll need a 32 bit KVM guest. Rich. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://et.redhat.com/~rjones/virt-df/

3 2

[libvirt] Web Interface -> LibVirt Communication?
by Henri Cook 09 Jul '08

09 Jul '08

Hi, I want to write a web tool that lets me administer my cluster from one location (including operations that things like virt-manager don't provide) - what's the proper way of achieving this? Can i communicate over TCP with libvirt on each server (is this documented?) or should I be trying to get my web app to auth and run virsh commands over SSH? Thanks. Henri

4 11

Re: [libvirt] libvirt-java
by Daniel Veillard 09 Jul '08

09 Jul '08

On Mon, Jul 07, 2008 at 01:14:40PM +0200, Tóth István wrote: > Daniel Veillard wrote: > > Ah, okay.... Well my perspective is that libvirt-java should include > > only parts which are directly releated to using libvirt on Java. I would > > really not try to reinvent or push any XML related APIs there, at least > > as part of libvirt bindings (I still have the scars from libxml2, believe > > me you don't want to push new XML APIs to programmers even specialized ones). > > > My thoughts exactly. Okay seems we agree on the fundamentals :-) > > But things like the XSD descriptiond might be useful generally, and > > IMHO are very tied to libvirt, so i think they are a good fit. > > > Absolutely. I've had major problems when I tried to use some Java XML > mapping libraries, because they can only process xsd, > and not the language you use. So having high-quality xsd definitions is > high on my personal wish-list. Okay > > Also fitting into this model are build makefiles for other environment, > > I understand that for most Java developpers, configure and make are really > > foreign tools, so adding ant/Eclipse/... build recipes for the package also > > makes a lot of sense. > > > Well, they certainly wouldn't hurt, but I don't see that as a priority. > Java-only programmers (end users) won't touch the JNI core parts. > They will just install the package, and get on with using it, by adding > the jar > to their devel enviroment, and perhaps the JNI lib file to the command line. > > These kind of users don't really care about the java-libvirt build > environment. okay, so just the Jar availability is sufficient for them, okidoc. > If you want to do something more interesting than renaming the java > classes and methods, > the you have to hit the JNI bindings, and that code is so ugly, that > no-one without some C experience will want to touch it. > > Having said that, I think that an ant build file would be the most > useful, as it can be used directly in all major > development enviroments. There is also this maven thing, that seems to > be widely used, but I still haven't figured out > exactly what that is :-) me either, though there are examples in the Fedora-java packaging pages. [...] > My current plans for java-libvirt are: > > 1. Add the storage API: It's really mostly just copy-paste-search > replace but it still takes some work okay > 2. There are some consistency problems with the naming of classes and > methods. I'd like to revisit the java api, and make some changes in > names, and maybe class structure Hum, better done early than late. Basically i would prefer to avoid pushing incompatible changes. what kind of inconsistencies problems ? > 3. There are many places where the C part leaks memory, this should also > be audited/ fixed. Ah, okay I will have to reread the bindings code then. Not sure how to track leaks, I doubt valgrind can work with java ... > Number 2 is what worries me, I don't know if it's a good idea to push > toFedora, when I know I want to make incompatible API changes soon. yes, which is why I would like to know a bit better :-) > (Or you can just say that you won't accept them, but I'm a big fan of > clean and consistent APIs, and the current one can be improved) > I believe that I will get around to doing 1. and 2. at least in late > july/early august, It's about a three day job, I just don't have that > three days right now :-( Maybe if you can expose what you think is wrong i can try to clean things up. Daniel -- Red Hat Virtualization group http://redhat.com/virtualization/ Daniel Veillard | virtualization library http://libvirt.org/ veillard(a)redhat.com | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/

3 3

[libvirt] PATCH: Fix ue of VIR_TEST_DEBUG var
by Daniel P. Berrange 09 Jul '08

09 Jul '08

The previous patch I did to add OOM testing accidentally disabled use of the VIR_TEST_DEBUG env variable when OOM testing is #ifdef'd out. This patch I just comitted makes sure its always used. Daniel diff -r d37150dc5895 tests/testutils.c --- a/tests/testutils.c Tue Jul 08 16:17:52 2008 +0100 +++ b/tests/testutils.c Wed Jul 09 11:23:27 2008 +0100 @@ -321,20 +321,24 @@ char **argv, int (*func)(int, char **)) { + char *debugStr; #if TEST_OOM int ret; int approxAlloc = 0; int n; - char *oomStr = NULL, *debugStr; + char *oomStr = NULL; int oomCount; int mp = 0; pid_t *workers; int worker = 0; +#endif + if ((debugStr = getenv("VIR_TEST_DEBUG")) != NULL) { if (virStrToLong_ui(debugStr, NULL, 10, &testDebug) < 0) testDebug = 0; } +#if TEST_OOM if ((oomStr = getenv("VIR_TEST_OOM")) != NULL) { if (virStrToLong_i(oomStr, NULL, 10, &oomCount) < 0) oomCount = 0; -- |: 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: Misc changes to policykit policy file
by Daniel P. Berrange 09 Jul '08

09 Jul '08

After discussions with policykit maintainers I've come to the conclusion that it is better for security if we default to 'auth_admin_keep_sesion' instead of 'auth_self_keep_session'. ie prompt for the root password (ala 'su') instead of the user's password (ala 'sudo'). This is because having access to libvirtd gives you very significant power over the host machine. Secondly, newer versions of policykit have imposed a naming constraint on policy files, so when we install our policy it needs to be in a file called org.libvirt.unix.policy, instead of just libvirt.policy. So there's a change to the Makefile to support this. Daniel diff -r e17c1bd119fc libvirt.spec.in --- a/libvirt.spec.in Wed Jul 09 09:26:26 2008 +0100 +++ b/libvirt.spec.in Wed Jul 09 10:32:24 2008 +0100 @@ -240,7 +240,7 @@ %dir %attr(0700, root, root) %{_localstatedir}/lib/libvirt/images/ %dir %attr(0700, root, root) %{_localstatedir}/lib/libvirt/boot/ %if %{with_polkit} -%{_datadir}/PolicyKit/policy/libvirtd.policy +%{_datadir}/PolicyKit/policy/org.libvirt.unix.policy %endif %dir %attr(0700, root, root) %{_localstatedir}/log/libvirt/qemu/ %if %{with_xen_proxy} diff -r e17c1bd119fc qemud/Makefile.am --- a/qemud/Makefile.am Wed Jul 09 09:26:26 2008 +0100 +++ b/qemud/Makefile.am Wed Jul 09 10:32:24 2008 +0100 @@ -13,7 +13,8 @@ remote_dispatch_proc_switch.h \ mdns.c mdns.h \ libvirtd.sasl \ - libvirtd.conf + libvirtd.conf \ + libvirtd.policy if RPCGEN SUFFIXES = .x @@ -75,9 +76,7 @@ if HAVE_POLKIT policydir = $(datadir)/PolicyKit/policy -policy_DATA = libvirtd.policy endif -EXTRA_DIST += libvirtd.policy if HAVE_AVAHI libvirtd_SOURCES += mdns.c mdns.h @@ -86,7 +85,7 @@ endif default_xml_dest = libvirt/qemu/networks/default.xml -install-data-local: install-init install-data-sasl +install-data-local: install-init install-data-sasl install-data-polkit mkdir -p $(DESTDIR)$(sysconfdir)/libvirt/qemu/networks/autostart $(INSTALL_DATA) $(srcdir)/default-network.xml \ $(DESTDIR)$(sysconfdir)/$(default_xml_dest) @@ -99,7 +98,7 @@ mkdir -p $(DESTDIR)$(localstatedir)/run/libvirt mkdir -p $(DESTDIR)$(localstatedir)/lib/libvirt -uninstall-local:: uninstall-init uninstall-data-sasl +uninstall-local:: uninstall-init uninstall-data-sasl install-data-polkit rm -f $(DESTDIR)$(sysconfdir)/libvirt/qemu/networks/autostart/default.xml rm -f $(DESTDIR)$(sysconfdir)/$(default_xml_dest) rmdir $(DESTDIR)$(sysconfdir)/libvirt/qemu/networks/autostart || : @@ -119,6 +118,18 @@ install-data-sasl: uninstall-data-sasl: endif + +if HAVE_POLKIT +install-data-polkit:: install-init + mkdir -p $(DESTDIR)$(policydir) + $(INSTALL_DATA) libvirtd.policy $(DESTDIR)$(policydir)/org.libvirt.unix.policy +uninstall-data-polkit:: install-init + rm -f $(DESTDIR)$(policydir)/org.libvirt.unix.policy +else +install-data-polkit:: +uninstall-data-polkit:: +endif + remote.c: remote_dispatch_prototypes.h \ remote_dispatch_localvars.h \ diff -r e17c1bd119fc qemud/libvirtd.policy --- a/qemud/libvirtd.policy Wed Jul 09 09:26:26 2008 +0100 +++ b/qemud/libvirtd.policy Wed Jul 09 10:32:24 2008 +0100 @@ -36,7 +36,7 @@ read-write mode for management, and we require user password --> <allow_any>no</allow_any> <allow_inactive>no</allow_inactive> - <allow_active>auth_self_keep_session</allow_active> + <allow_active>auth_admin_keep_session</allow_active> </defaults> </action> -</policyconfig> \ No newline at end of file +</policyconfig> -- |: 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] monitor cpu & network load
by Zu 09 Jul '08

09 Jul '08

Hi, I want to write a program to monitor the CPU and network usage of the Xen domUs. But I don't know how to implement. Can the libvirt APIs help me to do that? Does libvirt have the APIs to get the CPU and network usages? If yes I'll study libvirt APIs. Thanks Zu

3 2

[libvirt] PATCH: DOn't use policykit if running non-root
by Daniel P. Berrange 09 Jul '08

09 Jul '08

If the libvirtd daemon is running as non-root (ie a regular user) then there is no point in defaulting policykit, since only clients of same user account are able to connect to the UNIX socket anyway. This removes the pointless login prompt when using qemu:///session. Daniel diff -r 9a457b847ffc qemud/qemud.c --- a/qemud/qemud.c Tue Jul 08 16:17:52 2008 +0100 +++ b/qemud/qemud.c Wed Jul 09 09:24:53 2008 +0100 @@ -1926,6 +1926,14 @@ GET_CONF_STR (conf, filename, tcp_port); GET_CONF_STR (conf, filename, listen_addr); +#if HAVE_POLKIT + /* Change the default back to no auth for non-root */ + if (getuid() != 0 && auth_unix_rw == REMOTE_AUTH_POLKIT) + auth_unix_rw = REMOTE_AUTH_NONE; + if (getuid() != 0 && auth_unix_ro == REMOTE_AUTH_POLKIT) + auth_unix_ro = REMOTE_AUTH_NONE; +#endif + if (remoteConfigGetAuth(conf, "auth_unix_rw", &auth_unix_rw, filename) < 0) goto free_and_fail; #if HAVE_POLKIT -- |: 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] [PATCH] give a more useful diagnostic for tap-add failure w/ENOTSUP
by Jim Meyering 09 Jul '08

09 Jul '08

I invoked virt-isntall with --network=bridge:eth1 rather than --network=bridge:br1 and got the latter (current) diagnostic below. This change makes it so in this relatively common case people get a slightly more understandable diagnostic. >From 3d6dc8442fd9f1b0915c232040200154832fdf51 Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Tue, 8 Jul 2008 13:58:10 +0200 Subject: [PATCH] give a more useful diagnostic for tap-add failure w/ENOTSUP * src/qemu_conf.c (qemudNetworkIfaceConnect): Suggestion from Daniel P. Berrange. --- src/qemu_conf.c | 14 +++++++++++--- 1 files changed, 11 insertions(+), 3 deletions(-) diff --git a/src/qemu_conf.c b/src/qemu_conf.c index 17f0162..4b8c01e 100644 --- a/src/qemu_conf.c +++ b/src/qemu_conf.c @@ -2304,9 +2304,17 @@ qemudNetworkIfaceConnect(virConnectPtr conn, if ((err = brAddTap(driver->brctl, brname, ifname, BR_IFNAME_MAXLEN, &tapfd))) { - qemudReportError(conn, NULL, NULL, VIR_ERR_INTERNAL_ERROR, - _("Failed to add tap interface '%s' to bridge '%s' : %s"), - ifname, brname, strerror(err)); + if (errno == ENOTSUP) { + /* In this particular case, give a better diagnostic. */ + qemudReportError(conn, NULL, NULL, VIR_ERR_INTERNAL_ERROR, + _("Failed to add tap interface to bridge. " + "%s is not a bridge device"), brname); + } else { + qemudReportError(conn, NULL, NULL, VIR_ERR_INTERNAL_ERROR, + _("Failed to add tap interface '%s' " + "to bridge '%s' : %s"), + ifname, brname, strerror(err)); + } goto error; } -- 1.5.6.2.221.gb589

3 3

[libvirt] [PATCH] Missing htonl (etc) on MinGW
by Richard W.M. Jones 08 Jul '08

08 Jul '08

With the attached patch you can get all the way through a compile of libvirt using the MinGW cross-compiler. Basically MinGW lacks the htonl/ntohl/htons/ntohs functions. This simply adds the same functions from glibc (which has a compatible license) in a single file called "byteswap.h". The reason it worked previously was that PortableXDR used to provide these functions by linking to winsock2. However I have now removed that dependency from PortableXDR because it prevented a DLL being built of PortableXDR. This doesn't fully resolve dynamic linking of libvirt yet, but it is one step (or rather, one library) closer to happening. Rich. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones Read my OCaml programming blog: http://camltastic.blogspot.com/ Fedora now supports 59 OCaml packages (the OPEN alternative to F#) http://cocan.org/getting_started_with_ocaml_on_red_hat_and_fedora

2 2

[libvirt] incorrect VIR_DOMAIN_NONE usage
by Tóth István 08 Jul '08

08 Jul '08

As I was trying to understand exact semantics of the libvirt flags api, I found an error in the xs_internal.c file. When it wants to indicate that it cannot report the state of the domain, it user VIR_DOMAIN_NONE as a return value, which does not, in fact, refer to a domain state at all, but is a dummy flag for creating domains, instead of VIR_DOMAIN_NOSTATE. This patch does not affect the compiled code, only the readability. On a related note, the defined enum flags seem inconstent to me, half of them have explicitly named 0 default values, and half of them don't, it's a bit confusing. If enum virStorageVolDeleteFlags { VIR_STORAGE_VOL_DELETE_NORMAL = 0 : Delete metadata only (fast) VIR_STORAGE_VOL_DELETE_ZEROED = 1 : Clear all data to zeros (slow) } then why not enum virConnectFlags { VIR_CONNECT_RW = 0 : A read-write connection VIR_CONNECT_RO = 1 : A readonly connection } ? It would not affect existing code, and would make the library more consistent. regards István ? .project ? xs_internal_constant_semantics.patch Index: src/xs_internal.c =================================================================== RCS file: /data/cvs/libvirt/src/xs_internal.c,v retrieving revision 1.65 diff -u -p -r1.65 xs_internal.c --- src/xs_internal.c 10 Apr 2008 16:54:54 -0000 1.65 +++ src/xs_internal.c 1 Jul 2008 19:46:41 -0000 @@ -399,7 +399,7 @@ xenStoreGetDomainInfo(virDomainPtr domai info->state = VIR_DOMAIN_RUNNING; free(tmp); } else { - info->state = VIR_DOMAIN_NONE; + info->state = VIR_DOMAIN_NOSTATE; } tmp = virDomainDoStoreQuery(domain->conn, domain->id, "memory/target"); if (tmp != NULL) {

2 1

[Libvirt] [PATCH] 1+2=3 errors in docs
by Anton Protopopov 08 Jul '08

08 Jul '08

Hi. This patch fixes three examples given in docs/drvqemu.html.in and docs/ formatnetwork.html.in A.

2 1

[libvirt] libvirt and linux vserver
by ADNET Ghislain 08 Jul '08

08 Jul '08

Hi, i am new here , i looked on the web but cannot find anything about linux vserver support in libvirt. I found that one of the main dev of linux vserver submitted patches for the support but nothing seems to have been integrated from his work into libvirt. Do you have reviewed them ? Is libvirt a possibility to manage linux vservers ? * /From/: "Daniel Hokka Zakrisson" <daniel hozac com> * /To/: libvir-list redhat com * /Subject/: [Libvir] [PATCH] Linux-VServer support * /Date/: Wed, 9 Jan 2008 22:51:20 +0100 (CET) http://www.redhat.com/archives/libvir-list/2008-January/msg00097.html -- Cordialement, Ghislain

2 1