- Devel - libvirt List Archives

[libvirt] [PATCH] Detect bridge name collisions when defining a virtual network
by Cole Robinson 17 Feb '09

17 Feb '09

The attached patch is a partial fix for rhbz 479622. If a virtual network is defined with an explicit bridge name, make sure it doesn't collide with other virtual network bridge names. Thanks, Cole

2 1

[libvirt] [PATCH 1/4] fix problem with MKDIR_P vs mkdir_p
by Maximilian Wilhelm 17 Feb '09

17 Feb '09

From: Matthias Bolte <photon(a)upb.de> configure defines mkdir_p but src/Makefile.am uses MKDIR_P --- src/Makefile.am | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 9d934b4..eff9039 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -589,6 +589,8 @@ endif endif EXTRA_DIST += $(LXC_CONTROLLER_SOURCES) +MKDIR_P = $(mkdir_p) + install-exec-local: $(MKDIR_P) "$(DESTDIR)$(localstatedir)/cache/libvirt" $(MKDIR_P) "$(DESTDIR)$(localstatedir)/lib/libvirt/images" -- 1.5.6.5

3 14

[libvirt] Network Configuration question
by Remko Nolten 17 Feb '09

17 Feb '09

Hi! For a shared virtual hosting project with some friends we need a pretty specialized network configuration. Because we have virtually no time for experimenting (no pun intended), and the hosting organization has no experience with libvirt/KVM configurations they suggested to try and ask here. So, this is the situation: We have a server with one physical connection. Each user gets one virtual machine with its own IP-address (we have our own ip-range). The hosting organization gives us one "main" address 'A' with netmask 255.255.255.0 (/24) and a default gateway for our host machine. Besides this we get a number of addresses B1, B2, etc. with netmask 255.255.255.255 (/32) and without a default gateway. The switch just routes the traffic to our server for each ip-address we own (A and Bn). According to our hosting organization, the normal solution is to configure a loopback interface for each of the ip-addresses on the server so you can configure your software just to use one of the addresses. But we are not sure how this will work out in our virtual environment. Our intuition tells us we need to use the routing network forwarding mode, but when we bring the server to the data center we don't want to have any surprises. :) Here at home I don't have to possibility to create similar circumstances as in the data center due to some restriction from my ISP, so I hope someone can tell me if our intuition is correct or that we need to use some other method. Thanks in advance! -- Remko Nolten Tel: 06-45600767 E-mail: remko(a)nolten.nu

3 4

[libvirt] [PATCH] Don't check volume format for lvm backing store
by Cole Robinson 16 Feb '09

16 Feb '09

With the recently added COW support, we were trying to validate the passed logical volume's format, when there won't ever be one pulled from the xml (logical volumes don't have any valid format values). Just remove this check, and let any validation in this manner be done by the actual call to 'lvcreate'. Thanks, Cole

3 4

[libvirt] [PATCH] Fix generating python bindings
by Cole Robinson 16 Feb '09

16 Feb '09

The recent API additions virSaveLastError and virFreeError upset the python bindings generator. Add a rule to explicitly skip them for the bindings, since (as far as I can tell) the libvirtError class provides everything we need. Thanks, Cole

3 4

[libvirt] [PATCH] Fix virsh {net,pool}-edit
by Cole Robinson 16 Feb '09

16 Feb '09

The dumpxml commands for networks and pools don't support any flag arguments, and in fact explictly fail if flags != 0. This is not the case for vm dumpxml though, and flags were added to the base 'edit' implementation in virsh recently. The net and pool derivatives were not addressing this difference, so any net-edit or pool-edit attempt currently gives an error like: Network error : invalid argument in virNetworkGetXMLDesc The attached patch is one way to fix this. Thanks to Charles Duffy for the report. - Cole

3 7

[libvirt] Per-VM access control
by Jan Kasprzak 16 Feb '09

16 Feb '09

Hello, is it possible to run libvirt as a "hosting-like" environment? we would like to provide virtual machines for our users, but we would like them to be able to reset/reboot/poweroff only their own VMs, connect to the serial console of their own VMs only, and even maybe connect to the graphical console of their own VMs. Is it possible with libvirt, and without giving them shell access to the hosting machine? For the serial console, I thought about creating a ssh-only account with hardcoded "virsh console <their_machine_name>" command, but "virsh console" can be escaped from using ^], so this is not secure - this way they would be able to get access to other VMs as well. Thanks, -Yenya -- | Jan "Yenya" Kasprzak <kas at {fi.muni.cz - work | yenya.net - private}> | | GPG: ID 1024/D3498839 Fingerprint 0D99A7FB206605D7 8B35FCDE05B18A5E | | http://www.fi.muni.cz/~kas/ Journal: http://www.fi.muni.cz/~kas/blog/ | >> If you find yourself arguing with Alan Cox, you’re _probably_ wrong. << >> --James Morris in "How and Why You Should Become a Kernel Hacker" <<

3 2

[libvirt] Compilation Errors
by Russell 16 Feb '09

16 Feb '09

I'm trying to compile libvirt 0.6.0 on Gentoo, but keep encountering the following error: network_driver.c:72: error: expected specifier-qualifier-list before 'brControl' network_driver.c: In function 'networkFindActiveConfigs': <snip - just a bunch of 'struct network_driver' has no member named * > network_driver.c:1382: error: 'struct network_driver' has no member named 'networkConfigDir' network_driver.c:1384: error: 'struct network_driver' has no member named 'networkAutostartDir' network_driver.c:1390: error: 'struct network_driver' has no member named 'networkAutostartDir' network_driver.c:1391: error: 'struct network_driver' has no member named 'networkAutostartDir' make[3]: *** [network_driver.lo] Error 1 make[3]: Leaving directory `/home/russell/libvirt/libvirt-0.6.0/src' make[2]: *** [all] Error 2 make[2]: Leaving directory `/home/russell/libvirt/libvirt-0.6.0/src' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/russell/libvirt/libvirt-0.6.0' make: *** [all] Error 2 Every error following the first one refers to networkConfigDir, networkAutostartDir, or logDir, all listed after brControl in the definition of network_driver. What looks to me like the same error has been reported against 0.5.1 here (I haven't tried it myself): http://bugs.gentoo.org/show_bug.cgi?id=254568 My configuration options: ./configure --prefix=/usr --host=i686-pc-linux-gnu --mandir=/usr/share/man --infodir=/usr/share/info --datadir=/usr/share --sysconfdir=/etc --localstatedir=/var/lib --with-avahi --without-storage-iscsi --without-storage-lvm --without-lxc --with-hal --without-openvz --without-storage-disk --without-sasl --without-selinux --without-uml --without-xen --without-qemu --without-devkit --without-polkit --with-remote --disable-iptables-lokkit --localstatedir=/var --with-remote-pid-file=/var/run/libvirtd.pid --build=i686-pc-linux-gnu If other info would help resolve this I'd be happy to try and provide it. Thanks, -- Russell Haering System Administrator OSU Open Source Lab

3 4

[libvirt] [PATCH] Allow off-line removal of devices via xend
by john.levon＠sun.com 15 Feb '09

15 Feb '09

# HG changeset patch # User john.levon(a)sun.com # Date 1233765613 28800 # Node ID bda41ea0cbbdea409447686c30b7afb10b9cae85 # Parent e6b17082b9b9440ce14ad76bd2f4c003ae2404db Allow off-line removal of devices via xend We must avoid xenstore if the domain isn't running. For disks, this works in all xend versions; for NICs, however, this requires a change in netif.py to allow matching by MAC address. Preferring the device ID but falling back to the device ref works best with all versions of xend. Signed-off-by: John Levon <john.levon(a)sun.com> diff --git a/src/xend_internal.c b/src/xend_internal.c --- a/src/xend_internal.c +++ b/src/xend_internal.c @@ -92,11 +92,17 @@ xenDaemonFormatSxprNet(virConnectPtr con int xendConfigVersion, int isAttach); static int -virDomainXMLDevID(virDomainPtr domain, - virDomainDeviceDefPtr dev, - char *class, - char *ref, - int ref_len); +virDomainDevRef(virDomainPtr domain, + virDomainDeviceDefPtr dev, + char *class, + char *ref, + int ref_len); +static int +virDomainDevID(virDomainPtr domain, + int type, + const char *ref, + char *devid, + int devid_len); #endif #define virXendError(conn, code, fmt...) \ @@ -3857,7 +3863,7 @@ xenDaemonAttachDevice(virDomainPtr domai virDomainDeviceDefPtr dev = NULL; virDomainDefPtr def = NULL; virBuffer buf = VIR_BUFFER_INITIALIZER; - char class[8], ref[80]; + char class[8], ref[80], devid[80]; if ((domain == NULL) || (domain->conn == NULL) || (domain->name == NULL)) { virXendError((domain ? domain->conn : NULL), VIR_ERR_INVALID_ARG, @@ -3913,15 +3919,24 @@ xenDaemonAttachDevice(virDomainPtr domai sexpr = virBufferContentAndReset(&buf); - if (virDomainXMLDevID(domain, dev, class, ref, sizeof(ref))) { + devid[0] = '\0'; + + if (virDomainDevRef(domain, dev, class, ref, sizeof(ref)) == -1) + goto cleanup; + + if (domain->id > 0 && + virDomainDevID(domain, dev->type, ref, devid, sizeof(devid)) != -1) { + VIR_DEBUG("device_configure(ref = %s, devid = %s): %s", + ref, devid, sexpr); + /* device exists, attempt to modify it */ + ret = xend_op(domain->conn, domain->name, "op", "device_configure", + "config", sexpr, "dev", devid, NULL); + } else { + VIR_DEBUG("device_create(ref = %s, devid = %s): %s", + ref, devid, sexpr); /* device doesn't exist, define it */ ret = xend_op(domain->conn, domain->name, "op", "device_create", "config", sexpr, NULL); - } - else { - /* device exists, attempt to modify it */ - ret = xend_op(domain->conn, domain->name, "op", "device_configure", - "config", sexpr, "dev", ref, NULL); } cleanup: @@ -3944,7 +3959,7 @@ xenDaemonDetachDevice(virDomainPtr domai xenDaemonDetachDevice(virDomainPtr domain, const char *xml) { xenUnifiedPrivatePtr priv; - char class[8], ref[80]; + char class[8], ref[80], devid[80]; virDomainDeviceDefPtr dev = NULL; virDomainDefPtr def = NULL; int ret = -1; @@ -3975,11 +3990,40 @@ xenDaemonDetachDevice(virDomainPtr domai def, xml, VIR_DOMAIN_XML_INACTIVE))) goto cleanup; - if (virDomainXMLDevID(domain, dev, class, ref, sizeof(ref))) + if (dev->type != VIR_DOMAIN_DEVICE_NET && + dev->type != VIR_DOMAIN_DEVICE_DISK) { + virXendError(domain->conn, VIR_ERR_INTERNAL_ERROR, + _("can't detach device of type %d"), dev->type); goto cleanup; + } + + /* + * First acquire a static reference to the device (MAC address or + * destination disk). We'll use this if the domain isn't running, + * otherwise we'll rely upon the XenStore devid. + */ + + if (virDomainDevRef(domain, dev, class, ref, sizeof(ref)) == -1) { + virXendError(domain->conn, VIR_ERR_INTERNAL_ERROR, + _("can't detach device: not found ")); + goto cleanup; + } + + if (domain->id >= 0) { + if (virDomainDevID(domain, dev->type, ref, devid, sizeof(devid))) + virXendError(domain->conn, VIR_ERR_INTERNAL_ERROR, + _("can't detach device: not found ")); + goto cleanup; + } else { + strncpy(devid, ref, 80); + } + + VIR_DEBUG("device_destroy(type = %s, ref = %s, devid = %s)", + class, ref, devid); ret = xend_op(domain->conn, domain->name, "op", "device_destroy", - "type", class, "dev", ref, "force", "0", "rm_cfg", "1", NULL); + "type", class, "dev", devid, "force", "0", + "rm_cfg", "1", NULL); cleanup: virDomainDefFree(def); @@ -5501,12 +5545,48 @@ error: /** - * virDomainXMLDevID: + * virDomainXMLDevRef: * @domain: pointer to domain object * @dev: pointer to device config object * @class: Xen device class "vbd" or "vif" (OUT) * @ref: Xen device reference (OUT) * + * Return the class (vbd/vif) and reference (MAC address / disk + * destination) of the given device. + * + * Returns 0 in case of success, -1 in case of failure. + */ +static int +virDomainDevRef(virDomainPtr domain, + virDomainDeviceDefPtr dev, + char *class, + char *ref, + int ref_len) +{ + if (dev->type == VIR_DOMAIN_DEVICE_DISK) { + if (dev->data.disk->dst == NULL) + return -1; + strncpy(ref, dev->data.disk->dst, ref_len); + strcpy(class, "vbd"); + } else if (dev->type == VIR_DOMAIN_DEVICE_NET) { + virFormatMacAddr(dev->data.net->mac, ref); + strcpy(class, "vif"); + } else { + virXendError(NULL, VIR_ERR_NO_SUPPORT, + _("no support for device type %d"), dev->type); + return -1; + } + + return 0; +} + +/** + * virDomainXMLDevID: + * @domain: pointer to domain object + * @class: either 'vbd' or 'vif' + * @ref: Xen device reference + * @devid: Xen device ID (OUT) + * * Set class according to XML root, and: * - if disk, copy in ref the target name from description * - if network, get MAC address from description, scan XenStore and @@ -5515,54 +5595,41 @@ error: * Returns 0 in case of success, -1 in case of failure. */ static int -virDomainXMLDevID(virDomainPtr domain, - virDomainDeviceDefPtr dev, - char *class, - char *ref, - int ref_len) +virDomainDevID(virDomainPtr domain, + int type, + const char *ref, + char *devid, + int devid_len) { xenUnifiedPrivatePtr priv = domain->conn->privateData; - char *xref; - - if (dev->type == VIR_DOMAIN_DEVICE_DISK) { - strcpy(class, "vbd"); - if (dev->data.disk->dst == NULL) - return -1; - xenUnifiedLock(priv); - xref = xenStoreDomainGetDiskID(domain->conn, domain->id, - dev->data.disk->dst); - xenUnifiedUnlock(priv); - if (xref == NULL) - return -1; - - strncpy(ref, xref, ref_len); - free(xref); - ref[ref_len - 1] = '\0'; - } else if (dev->type == VIR_DOMAIN_DEVICE_NET) { - char mac[30]; - virDomainNetDefPtr def = dev->data.net; - snprintf(mac, sizeof(mac), "%02x:%02x:%02x:%02x:%02x:%02x", - def->mac[0], def->mac[1], def->mac[2], - def->mac[3], def->mac[4], def->mac[5]); - - strcpy(class, "vif"); - - xenUnifiedLock(priv); - xref = xenStoreDomainGetNetworkID(domain->conn, domain->id, - mac); - xenUnifiedUnlock(priv); - if (xref == NULL) - return -1; - - strncpy(ref, xref, ref_len); - free(xref); - ref[ref_len - 1] = '\0'; + char *id = NULL; + + if (domain->id < 0) { + virXendError(domain->conn, VIR_ERR_INTERNAL_ERROR, + _("devid not available for shutdown domain")); + return -1; + } + + xenUnifiedLock(priv); + + if (type == VIR_DOMAIN_DEVICE_DISK) { + id = xenStoreDomainGetDiskID(domain->conn, domain->id, ref); + + } else if (type == VIR_DOMAIN_DEVICE_NET) { + id = xenStoreDomainGetNetworkID(domain->conn, domain->id, ref); } else { virXendError(NULL, VIR_ERR_NO_SUPPORT, - "%s", _("hotplug of device type not supported")); + _("no support for device type %d"), type); + } + + xenUnifiedUnlock(priv); + + if (id == NULL) return -1; - } - + + strncpy(devid, id, devid_len); + VIR_FREE(id); + devid[devid_len - 1] = '\0'; return 0; }

7 13

[libvirt] Libvirt /proc/cpuinfo processing fails on Linux/Sparc
by Richard W.M. Jones 14 Feb '09

14 Feb '09

Below is what /proc/cpuinfo looks like on an Ultrasparc T1 running Linux. I haven't looked into it in detail yet, but the report from Dennis Gilmore is that this breaks every libvirt command. Presumably we are parsing /proc/cpuinfo at libvirt startup ... (Initially reported by Dennis Gilmore) Rich. cpu : UltraSparc T1 (Niagara) fpu : UltraSparc T1 integrated FPU prom : OBP 4.29.0.a 2008/09/15 11:59 type : sun4v ncpus probed : 32 ncpus active : 32 D$ parity tl1 : 0 I$ parity tl1 : 0 Cpu0ClkTck : 000000003b9aca00 Cpu1ClkTck : 000000003b9aca00 Cpu2ClkTck : 000000003b9aca00 Cpu3ClkTck : 000000003b9aca00 Cpu4ClkTck : 000000003b9aca00 Cpu5ClkTck : 000000003b9aca00 Cpu6ClkTck : 000000003b9aca00 Cpu7ClkTck : 000000003b9aca00 Cpu8ClkTck : 000000003b9aca00 Cpu9ClkTck : 000000003b9aca00 Cpu10ClkTck : 000000003b9aca00 Cpu11ClkTck : 000000003b9aca00 Cpu12ClkTck : 000000003b9aca00 Cpu13ClkTck : 000000003b9aca00 Cpu14ClkTck : 000000003b9aca00 Cpu15ClkTck : 000000003b9aca00 Cpu16ClkTck : 000000003b9aca00 Cpu17ClkTck : 000000003b9aca00 Cpu18ClkTck : 000000003b9aca00 Cpu19ClkTck : 000000003b9aca00 Cpu20ClkTck : 000000003b9aca00 Cpu21ClkTck : 000000003b9aca00 Cpu22ClkTck : 000000003b9aca00 Cpu23ClkTck : 000000003b9aca00 Cpu24ClkTck : 000000003b9aca00 Cpu25ClkTck : 000000003b9aca00 Cpu26ClkTck : 000000003b9aca00 Cpu27ClkTck : 000000003b9aca00 Cpu28ClkTck : 000000003b9aca00 Cpu29ClkTck : 000000003b9aca00 Cpu30ClkTck : 000000003b9aca00 Cpu31ClkTck : 000000003b9aca00 MMU Type : Hypervisor (sun4v) State: CPU0: online CPU1: online CPU2: online CPU3: online CPU4: online CPU5: online CPU6: online CPU7: online CPU8: online CPU9: online CPU10: online CPU11: online CPU12: online CPU13: online CPU14: online CPU15: online CPU16: online CPU17: online CPU18: online CPU19: online CPU20: online CPU21: online CPU22: online CPU23: online CPU24: online CPU25: online CPU26: online CPU27: online CPU28: online CPU29: online CPU30: online CPU31: online -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://et.redhat.com/~rjones/virt-top

2 6