- Users - libvirt List Archives

[libvirt-users] save - restore problem
by Fırat KÜÇÜK 18 Mar '16

18 Mar '16

I was trying to live-migrate a VM. I have not succeeded for now. Instead live-migration Just tried simple save and restore action in same host. virsh save my-domain my-domain.saved virsh restore my-domain.saved Everything seems fine after restoring. But after 5 minutes VM halted. No network connectivity. No response to VNC keystrokes. Totally halted. Is there any suggestion?

1 1

[libvirt-users] Fwd: [Issue]: Regarding client socket getting closed from the server once the lxc container is started
by rammohan madhusudan 18 Mar '16

18 Mar '16

Hi Folks, Using libvirt python bindings we are creating an lxc container.Here is the problem that we see sometimes (say 20 % of the time) when we create a new container. 1.container gets created, and also starts.However the we are not able to enter the namepace of the container.It throws an error initPid not available.We see that the using netstat command , socket connection is closed. 2.To get around this problem we have to stop and start the container again.We see that socket under (/var/run/libvirt/*) connection is established and we are able to enter the namespace. Enabled the libvirtd debug logs to debug this issue. For *success* case we see that for new client connection gets created and is able to handle async incoming events, *2016-03-12 08:18:55.748+0000: 1247: info : virObjectNew:202 : OBJECT_NEW: obj=0x7fed54005460 classname=virLXCMonitor* *2016-03-12 08:18:55.748+0000: 1247: debug : virNetSocketNew:159 : localAddr=0x7fed7cd1d170 remoteAddr=0x7fed7cd1d200 fd=28 errfd=-1 pid=0* *2016-03-12 08:18:55.749+0000: 1247: info : virObjectNew:202 : OBJECT_NEW: obj=0x7fed54009040 classname=virNetSocket* *2016-03-12 08:18:55.749+0000: 1247: info : virNetSocketNew:209 : RPC_SOCKET_NEW: sock=0x7fed54009040 fd=28 errfd=-1 pid=0 localAddr=127.0.0.1;0, remoteAddr=127.0.0.1;0* *2016-03-12 08:18:55.749+0000: 1247: info : virObjectNew:202 : OBJECT_NEW: obj=0x7fed54009d10 classname=virNetClient* *2016-03-12 08:18:55.749+0000: 1247: info : virNetClientNew:327 : RPC_CLIENT_NEW: client=0x7fed54009d10 sock=0x7fed54009040* *2016-03-12 08:18:55.749+0000: 1247: info : virObjectRef:296 : OBJECT_REF: obj=0x7fed54009d10* *2016-03-12 08:18:55.749+0000: 1247: info : virObjectRef:296 : OBJECT_REF: obj=0x7fed54009040* *2016-03-12 08:18:55.750+0000: 1247: info : virObjectNew:202 : OBJECT_NEW: obj=0x7fed540009a0 classname=virNetClientProgram* *2016-03-12 08:18:55.750+0000: 1247: info : virObjectRef:296 : OBJECT_REF: obj=0x7fed540009a0* *2016-03-12 08:18:55.750+0000: 1247: info : virObjectRef:296 : OBJECT_REF: obj=0x7fed54005460* *2016-03-12 08:18:55.750+0000: 1247: info : virObjectUnref:259 : OBJECT_UNREF: obj=0x7fed5c168eb0* *2016-03-12 08:18:55.750+0000: 1247: debug : virLXCProcessCleanInterfaces:475 : Cleared net names: eth0 * *2016-03-12 08:18:55.750+0000: 1247: info : virObjectUnref:259 : OBJECT_UNREF: obj=0x7fed5c168eb0* *2016-03-12 08:18:55.750+0000: 1247: info : virObjectUnref:259 : OBJECT_UNREF: obj=0x7fed5c169600* *2016-03-12 08:18:55.755+0000: 1244: debug : virNetClientIncomingEvent:1808 : client=0x7fed54009d10 wantclose=0* *2016-03-12 08:18:55.755+0000: 1244: debug : virNetClientIncomingEvent:1816 : Event fired 0x7fed54009040 1* *2016-03-12 08:18:55.755+0000: 1244: debug : virNetMessageDecodeLength:151 : Got length, now need 36 total (32 more)* *2016-03-12 08:18:55.756+0000: 1244: info : virNetClientCallDispatch:1116 : RPC_CLIENT_MSG_RX: client=0x7fed54009d10 len=36 prog=305402420 vers=1 proc=2 type=2 status=0 serial=1* *2016-03-12 08:18:55.756+0000: 1244: debug : virKeepAliveCheckMessage:377 : ka=(nil), client=0x7fed81fc5ed4, msg=0x7fed54009d78* *2016-03-12 08:18:55.756+0000: 1244: debug : virNetClientProgramDispatch:220 : prog=305402420 ver=1 type=2 status=0 serial=1 proc=2* *2016-03-12 08:18:55.756+0000: 1244: debug : virLXCMonitorHandleEventInit:109 : Event init 1420 * For *failure* case ,we see that the client socket connection is initiated and gets closed immediately after receiving an incoming event.In this case, I don’t see an object for virNetClientProgram being created. Incoming event comes in and since the its unable to find client->prog it bails out and closes the connection. Snaphost of the code, static int virNetClientCallDispatchMessage(virNetClientPtr client) { size_t i; virNetClientProgramPtr prog = NULL; for (i = 0; i < client->nprograms; i++) { if (virNetClientProgramMatches(client->programs[i], &client->msg)) { prog = client->programs[i]; break; } } if (!prog) { * VIR_DEBUG("No program found for event with prog=%d vers=%d",* * client->msg.header.prog, client->msg.header.vers);* return -1; } *2016-03-12 08:19:15.935+0000: 1246: info : virObjectRef:296 : OBJECT_REF: obj=0x7fed5c168eb0* *2016-03-12 08:19:15.935+0000: 1246: info : virObjectRef:296 : OBJECT_REF: obj=0x7fed82bd7bc0* *2016-03-12 08:19:15.935+0000: 1246: info : virObjectNew:202 : OBJECT_NEW: obj=0x7fed82bd8120 classname=virLXCMonitor* *2016-03-12 08:19:15.935+0000: 1246: debug : virNetSocketNew:159 : localAddr=0x7fed7d51e170 remoteAddr=0x7fed7d51e200 fd=31 errfd=-1 pid=0* *2016-03-12 08:19:15.936+0000: 1246: info : virObjectNew:202 : OBJECT_NEW: obj=0x7fed82bd8660 classname=virNetSocket* *2016-03-12 08:19:15.936+0000: 1246: info : virNetSocketNew:209 : RPC_SOCKET_NEW: sock=0x7fed82bd8660 fd=31 errfd=-1 pid=0 localAddr=127.0.0.1;0, remoteAddr=127.0.0.1;0* *2016-03-12 08:19:15.936+0000: 1246: info : virObjectNew:202 : OBJECT_NEW: obj=0x7fed82bd8ca0 classname=virNetClient* *2016-03-12 08:19:15.936+0000: 1246: info : virNetClientNew:327 : RPC_CLIENT_NEW: client=0x7fed82bd8ca0 sock=0x7fed82bd8660* *2016-03-12 08:19:15.936+0000: 1246: info : virObjectRef:296 : OBJECT_REF: obj=0x7fed82bd8ca0* *2016-03-12 08:19:15.936+0000: 1246: info : virObjectRef:296 : OBJECT_REF: obj=0x7fed82bd8660* *2016-03-12 08:19:15.942+0000: 1244: debug : virNetClientIncomingEvent:1808 : client=0x7fed82bd8ca0 wantclose=0* *2016-03-12 08:19:15.942+0000: 1244: debug : virNetClientIncomingEvent:1816 : Event fired 0x7fed82bd8660 1* *2016-03-12 08:19:15.942+0000: 1244: debug : virNetMessageDecodeLength:151 : Got length, now need 36 total (32 more)* *2016-03-12 08:19:15.942+0000: 1244: info : virNetClientCallDispatch:1116 : RPC_CLIENT_MSG_RX: client=0x7fed82bd8ca0 len=36 prog=305402420 vers=1 proc=2 type=2 status=0 serial=1* *2016-03-12 08:19:15.942+0000: 1244: debug : virKeepAliveCheckMessage:377 : ka=(nil), client=0x7fed81fc5ed4, msg=0x7fed82bd8d08* *2016-03-12 08:19:15.942+0000: 1244: debug : virNetClientCallDispatchMessage:1008 : No program found for event with prog=305402420 vers=1* *2016-03-12 08:19:15.942+0000: 1244: debug : virNetMessageClear:57 : msg=0x7fed82bd8d08 nfds=0* *2016-03-12 08:19:15.942+0000: 1244: debug : virNetClientMarkClose:632 : client=0x7fed82bd8ca0, reason=0* *2016-03-12 08:19:15.942+0000: 1244: debug : virNetClientCloseLocked:648 : client=0x7fed82bd8ca0, sock=0x7fed82bd8660, reason=0* Here is the snapshot of code , virLXCMonitorPtr virLXCMonitorNew(virDomainObjPtr vm, const char *socketdir, virLXCMonitorCallbacksPtr cb) { virLXCMonitorPtr mon; char *sockpath = NULL; if (virLXCMonitorInitialize() < 0) return NULL; if (!(mon = virObjectLockableNew(virLXCMonitorClass))) return NULL; if (virAsprintf(&sockpath, "%s/%s.sock", socketdir, vm->def->name) < 0) goto error; if (!(mon->client = virNetClientNewUNIX(sockpath, false, NULL))) goto error; * if (virNetClientRegisterAsyncIO(mon->client) < 0)* * goto error;* * if (!(mon->program = virNetClientProgramNew(VIR_LXC_MONITOR_PROGRAM,* * VIR_LXC_MONITOR_PROGRAM_VERSION,* * virLXCMonitorEvents,* * ARRAY_CARDINALITY(virLXCMonitorEvents),* * mon)))* * goto error;* * if (virNetClientAddProgram(mon->client,* * mon->program) < 0)* * goto error;* * mon->vm = vm;* * memcpy(&mon->cb, cb, sizeof(mon->cb));* virObjectRef(mon); virNetClientSetCloseCallback(mon->client, virLXCMonitorEOFNotify, mon, virLXCMonitorCloseFreeCallback); Is the problem occurring due to invocation of “*virNetClientRegisterAsyncIO" api before the virNetClientAddProgram.Probably once we register for aysnc IO , immediately an event comes in and that thread takes priority and bails out since it does not find the client->prog?Also the client is not retrying to establish a new connection.* *Please let me any thoughts/comments.Is there any patch already available which has fixed this issue?We are using libvirt 1.2.15* *-Thanks,* *Rammohan*

1 0

[libvirt-users] Questions regarding hostdev scsi
by Martin Polednik 18 Mar '16

18 Mar '16

Hi! I'm oVirt developer responsible for most of 'hostdev' support. While working on SCSI passthrough (that is hostdev type='scsi'), I've encountered few issues I'm not sure how to solve somewhat effectively and nicely. Just a note - oVirt by default disables 'dynamic_ownership', meaning we have to handle endpoint ownership/labeling ourselves. This is not something I can change in a short term. Also, oVirt uses libvirt's python API, I'll do my best using the original C names. To report and construct the hostdev element, I am using virConnectListAllNodeDevices. To get information about the devices, virNodeDeviceGetXMLDesc is called on each device. For PCI and USB devices, XML of the device contains everything needed to a) construct the element, b) fix endpoint permissions. SCSI device becomes more difficult as the information is scattered between multiple devices. Devices I have encountered contain this subtree: <device> <name>scsi_host4</name> <path>/sys/devices/pci0000:00/0000:00:1f.2/ata5/host4</path> <parent>pci_0000_00_1f_2</parent> <capability type='scsi_host'> <host>4</host> <unique_id>5</unique_id> </capability> </device> <device> <name>scsi_target4_0_0</name> <path>/sys/devices/pci0000:00/0000:00:1f.2/ata5/host4/target4:0:0</path> <parent>scsi_host4</parent> <capability type='scsi_target'> <target>target4:0:0</target> </capability> </device> <device> <name>scsi_4_0_0_0</name> <path>/sys/devices/pci0000:00/0000:00:1f.2/ata5/host4/target4:0:0/4:0:0:0</path> <parent>scsi_target4_0_0</parent> <driver> <name>sd</name> </driver> <capability type='scsi'> <host>4</host> <bus>0</bus> <target>0</target> <lun>0</lun> <type>disk</type> </capability> </device> <device> <name>block_sdb_Samsung_SSD_850_PRO_256GB_S251NXAGB42213R</name> <path>/sys/devices/pci0000:00/0000:00:1f.2/ata5/host4/target4:0:0/4:0:0:0/block/sdb</path> <parent>scsi_4_0_0_0</parent> <capability type='storage'> <block>/dev/sdb</block> <bus>ata</bus> <drive_type>disk</drive_type> <model>Samsung SSD 850</model> <vendor>ATA</vendor> <serial>Samsung_SSD_850_PRO_256GB_S251NXAGB42213R</serial> <size>256060514304</size> <logical_block_size>512</logical_block_size> <num_blocks>500118192</num_blocks> </capability> </device> <device> <name>scsi_generic_sg1</name> <path>/sys/devices/pci0000:00/0000:00:1f.2/ata5/host4/target4:0:0/4:0:0:0/scsi_generic/sg1</path> <parent>scsi_4_0_0_0</parent> <capability type='scsi_generic'> <char>/dev/sg1</char> </capability> </device> To construct the element, information from device scsi_4_0_0_0 is needed for <address> element (minus the host element). Adapter element is one of the places that I am not sure which information I can rely on, currently 'host' element from scsi_4_0_0_0 is used to form string 'scsi_host{host}'. Is this correct? Reliable? Whole different issue is locating the endpoint (/dev/sg{int}) to set the permissions. From the subtree, it's apparent that it is included in device called scsi_generic_sg1, but there is no *direct* link between scsi_4_0_0_0 and scsi_generic_sg1. At this point, we present user with devices matching capability='scsi', therefore to get the information 2 additional parses of the tree would have to be done. Other way would be reporting 'scsi_generic' capability, where only 2 additional virNodeDeviceLookupByName calls would be required (to descend 2 levels). Is there anything *really* wrong with this approach? Are there any other options without dynamic ownershit? Seclabel doesn't seem to be available. I'm thankful for any hints. Regards, mpolednik

2 1

[libvirt-users] Backup with external snapshots
by Dominique Ramaekers 18 Mar '16

18 Mar '16

Hi everybody, Following this user mailing list over a period of time, I saw a few messages regarding not being able using the external snapshot feature. I refer to this procedure: http://wiki.libvirt.org/page/Live-disk-backup-with-active-blockcommit I would like to mention this hoping it will help those users: Using external snapshots works perfect again in qemu 2.5.0 together with this patch https://bugzilla.redhat.com/show_bug.cgi?id=1300209 Full version list on which I did my testing: Compiled against library: libvirt 1.3.1 Using library: libvirt 1.3.1 Gebruikte API: QEMU 1.3.1 Draaiende hypervisor: QEMU 2.5.0 Greetings, Dominique.

1 0

[libvirt-users] share nothing live migration causes read only filesystem
by Fırat KÜÇÜK 17 Mar '16

17 Mar '16

Hello, I started live migration job for a guest with the following command: virsh migrate --live --domain my-guest --verbose --persistent --copy-storage-all --desturi qemu+ssh://root@my-otherhost/system Guest system disk cache mode is none: <disk type='file' device='disk'> <driver name='qemu' type='qcow2' cache='none'/> <source file='/opt/datastore/bulutkare-1118.img'/> <backingStore/> <target dev='vda' bus='virtio'/> <alias name='virtio-disk0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </disk> after live migration completed, filesystem blocked write operations. root@my-guest:~/tmp# touch hello touch: cannot touch ‘hello’: Read-only file system What can be the reason? What is the best of migrating a VM to a target machine? Additional info: Guest operating system is Ubuntu 14.04 x64 Source/Target Host operating system is Ubuntu 15.04 x64 Guest file system is ext4 + swap Source/Target Host file system is XFS Source/Target Qemu version 1.2.2 Source/Target libvirt 1.2.12 This is from target libvirt log. It can be important: /build/qemu-Bjjgxn/qemu-2.2+dfsg/nbd.c:nbd_co_receive_request():L1146: reading from socket failed

2 1

[libvirt-users] disabling disk cache
by Fırat KÜÇÜK 17 Mar '16

17 Mar '16

Hello, is it possible to disable disk (vda) cache without rebooting system? Thanks.

2 1

[libvirt-users] Re：Re: [help] How to modify the vgamem value ?
by hongzhen_luo＠sina.com 16 Mar '16

16 Mar '16

Dear Mr Holger and Mr Pavel Thanks for your help , I am going to try to modify the vgamem value . ^ ^Have a nice day ... Best regards >Hi Pavel, >I have to apologize. After your post I study the page again and I >found your descrption. >Best regards, >Am 14.03.2016 um 10:20 schrieb Pavel Hrdina: > On Mon, Mar 14, 2016 at 09:41:36AM +0100, Holger Schranz wrote: >> Hi Luo, >> >> the vgamem is limited inside. Myself, I use QEMU 2.5/Virgil3D to expand >> the vgamem. > > It's not limited and can be modified inside the xml, the only limitation is that > the value has to be a power of 2 and for QXL the value must be at least 1024. > There is an exception, qemu can and in some cases updates the value internally, > libvirt after the guest is started will get the actual value from qemu and > update live XML. > > Pavel >

1 0

[libvirt-users] Unsupported network type ethernet using LXC
by abhishek jain 15 Mar '16

15 Mar '16

Hi Team I'm trying to launch VM using LXC as a hypervisor with interace="ethernet" configuration and I'm getting below errors.. *virsh -c lxc: start instance-00000321error: Failed to start domain instance-00000321error: internal error: Unsupported network type ethernet* Below is the xml file which I'm using .... cat right.xml <domain type='lxc'> <name>instance-00000321</name> <uuid>13dd822c-a99d-4f28-9f4e-53dfb744efcc</uuid> <memory unit='KiB'>524288</memory> <currentMemory unit='KiB'>524288</currentMemory> <vcpu placement='static'>1</vcpu> <resource> <partition>/machine</partition> </resource> <os> <type arch='aarch64' machine='virt'>exe</type> <init>/bin/sh</init> <cmdline>root=/dev/ram0 rw console=ttyAMA0 rootwait earlyprintk</cmdline> </os> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <emulator>/usr/lib/libvirt/libvirt_lxc</emulator> <filesystem type='mount' accessmode='passthrough'> <source dir='/'/> <target dir='/'/> </filesystem> <interface type='ethernet'> <mac address='02:74:ff:9d:52:f5'/> <script path=''/> <target dev='tap74ff9d52-f5'/> </interface> <interface type='ethernet'> <mac address='02:9b:a6:1e:67:5b'/> <script path=''/> <target dev='tap9ba61e67-5b'/> </interface> <console type='pty'> <target type='lxc' port='0'/> </console> </devices> <seclabel type='none'/> </domain> However I'm able to launch the same above xml file using KVM as hypervisor on the same host. Please help me regarding this. --Thanks Abhishek Jain

1 1

[libvirt-users] TCP Tunnel Info
by bancfc＠openmailbox.org 14 Mar '16

14 Mar '16

libvirt-users(a)redhat.com TCP Tunnel Info I've been looking at the TCP Tunnel network feature as a potential replacement for the extra private internal networking configuration file. Usecase: This network is supposed to go from VM1 to VM2 without DHCP, DNS or any incoming/outgoing connections to the host or outside world possible. https://libvirt.org/formatdomain.html#elementsNICSTCP * To make sure I understand, adding the TCP Tunnel setting for both VMs 1 and 2 is enough to do what I need? (force them to exclusively communicate without the need for adding a new network as typically done). * If another set of VMs 3 and 4 are running and connected to each other but I want to make sure they cannot connect to VMs 1 and 2, what source addresses should be used to isolate these 2 networks? Do you follow CIDR rules? * For example if the chosen source address is 10.152.152.11 for VMs 1 and 2 what should the other network have? * Going more complicated. Can one VM participate in two separate TCP Tunnel networks while keeping them isolated? Topology: VM1(virtual NIC1) <-> VM2 | VM1(virtual NIC2) <-> VM3 VM2 and 3 can only talk to VM1 but not to each other in this example. * Offtopic: Do your answers similarly apply for using the other Multicast and UDP options too? I can explain better if I'm not making any sense.

2 1

[libvirt-users] [help] How to modify the vgamem value ?
by hongzhen_luo＠sina.com 14 Mar '16

14 Mar '16

Dear everyone Do you know how to modify vgamem value in the .xml file ? such as :<video> <model type='qxl' ram='65536' vram='65536' vgamem='16384' heads='1'/> // from instance-0000015e.xml <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/</video> I plan to increase the volume of vgamem , but it seems that always keep the original value ‘16384’. Was it limited by QEMU? or others？ Please tell me how I can I modify it ? it's better that help me to make a example for modifing the vgamem..... Thank you very much Best Regards .

3 3