May 2017 - Users - libvirt List Archives

[libvirt-users] ceph rbd pool and libvirt manageability (virt-install)
by Jelle de Jong 29 Aug '18

29 Aug '18

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello everybody, I created a rbd pool and activated it, but I can't seem to create volumes in it with virsh or virt-install? # virsh pool-dumpxml myrbdpool <pool type='rbd'> <name>myrbdpool</name> <uuid>2d786f7a-2df3-4d79-ae60-1535bcf1c6b5</uuid> <capacity unit='bytes'>6997998301184</capacity> <allocation unit='bytes'>10309227031</allocation> <available unit='bytes'>6977204658176</available> <source> <host name='ceph01.powercraft.nl' port='6789'/> <host name='ceph02.powercraft.nl' port='6789'/> <host name='ceph03.powercraft.nl' port='6789'/> <name>libvirt-pool</name> <auth type='ceph' username='libvirt'> <secret uuid='029a334e-ed57-4293-bb99-ffafa8867122'/> </auth> </source> </pool> # virt-install --version 1.0.1 # virsh --version 1.2.9 I ended using virsh edit ceph-test.powercraft.nl and making creating the disk manually. <disk type='network' device='disk'> <auth username='libvirt'> <secret type='ceph' uuid='029a334e-ed57-4293-bb99-ffafa8867122'/> </auth> <source protocol='rbd' name='libvirt-pool/kvm01-storage'> <host name='ceph01.powercraft.nl' port='6789'/> <host name='ceph02.powercraft.nl' port='6789'/> <host name='ceph03.powercraft.nl' port='6789'/> </source> <target dev='vdc' bus='virtio'/> </disk> I use virt-install a lot to define, import and undefine domains, how can I use virt-install to manage my rdb disks? Kind regards, Jelle de Jong -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iJwEAQECAAYFAlV1xlQACgkQ1WclBW9j5HkbPQP+PjNrzvlqysslOp2Yk7wH4Mxy 2sh2dn96G0KOAHEeEn3BN6IWlnD1TADZbHdpTtMwkdv48Xwn0sP1s+3QDM4pb3gP n+z+dVxS8FouDIy/eiso3IBCj3g4TWbEX8ZHqs3jKqe0lZgAXBzB9xYSUowcEBrZ ddkPbr8p8ozWyOG+9V8= =lkK7 -----END PGP SIGNATURE-----

2 3

[libvirt-users] virRandomBits - not very random
by Brian Rak 01 Jun '18

01 Jun '18

I just ran into an issue where I had about 30 guests get duplicate mac addresses assigned. These were scattered across 30 different machines. Some debugging revealed that: 1) All the host machines were restarted within a couple seconds of each other 2) All the host machines had fairly similar libvirtd pids (within ~100 PIDs of each other) 3) Libvirt seeds the RNG using 'time(NULL) ^ getpid()' This perfectly explains why I saw so many duplicate mac addresses. Why is the RNG seed such a predictable value? Surely there has to be a better source of a random seed then the timestamp and the pid? The PID seems to me to be a very bad source of any randomness. I just ran a test across 60 of our hosts. 43 of them shared their PID with at least one other machine.

7 13

[libvirt-users] No way to stop virStream error after guest stop
by JosÉ Luis Valencia Gutierrez 02 Jun '17

02 Jun '17

Hello everyone, I am opening a virChannel(unix) to a domain and receiving data with a non-blocking virStream using events, when the connected domain gets stopped(which deletes the channel unix socket) by calling destroy, shutdown, pause or migrate on that domain, with the stream open the read event is triggered repeatedly, and virStreamRecv returns 0 bytes indicating EOF but neither virStreamFinish nor virStreamFinish is working to stop the stream to trigger the event . Each time the event is called I got this errors. libvirt: I/O Stream Utils error : this function is not supported by the connection driver: virStreamRecv libvirt: I/O Stream Utils error : this function is not supported by the connection driver: virStreamFinish libvirt: I/O Stream Utils error : this function is not supported by the connection driver: virStreamEventRemoveCallback Is there other way to stop getting this errors? or perhaps this is a bug. Thanks in advance. Jose Valencia

2 8

[libvirt-users] Hopefully simple question
by Leroy Tennison 01 Jun '17

01 Jun '17

Is there a way to determine what the hypervisor default NIC is? I'm using kvm/libvirt/qemu. If this is the wrong forum please please reply with that.

3 2

[libvirt-users] Another hopefully simple question
by Leroy Tennison 01 Jun '17

01 Jun '17

Actually the last for now. Is there a way to determine from the hypervisor (kvm/libvirt/qemu) if a VM has unmounted a partition?

2 1

[libvirt-users] can't establish more than 1000 connections with virsh
by dw 26 May '17

26 May '17

Hi: I'm trying to connect with libvirtd with virsh from a remote PC,but only can establish 1000 connections. If try more connections,prompt: "error: failed to connect to the hypervisor error: Failed to open file '/etc/libvirt/libvirt.conf': Too many open files" I tried in another PC,also get the same prompt. Anybody know why? Thanks! environment: libvirt:libvirt-2.0.0-10.el7 OS:centos7.3

2 6

[libvirt-users] nwfilter - how to protect VM's
by Thiago Oliveira 25 May '17

25 May '17

Hello guys! I have just one physical NIC (eth0) and some libvirt networks. Each network created represents one customer and it is working fine. My question is how can I protect between the customer's networks? For example: The host 192.168.2.2 from the network 192.168.2.0/30 can not access any VM in others networks but the 192.168.2.2 can access 192.168.0.1 because it is the gateway to the internet and this host can receive incoming connections from the internet. I would like to tag with VLAN for example but I don't have sure if this is the best way. See below my route table [image: Imagem inline 1] Thank you guys. Thiago

1 0

[libvirt-users] How to show underlying QEMU command line using virsh
by Roman Storozhenko 25 May '17

25 May '17

Hello, everybody I use virt to run my VM's and QEMU as a hypervisor. There is a file describing my virtual machine: /etc/libvirt/qemu/centos7.0.xml As you know it contains xml representation of my VM. Is there any way to show underlying qemu command that virsh runs using that file? For example: qemu-system-x86-64 -m 1024 -smp 2 -... If no, is there any way to figure it out during VM run ? --- Thanks in advance, Roman Storozhenko

4 3

[libvirt-users] How to set guest console size in libvirt configuration options?
by Roman Storozhenko 25 May '17

25 May '17

Hello, everybody I use virsh to run my VM and QEMU as a hypervisor. When I connect to it, it shows me a console with a some predefined size. I know that I am able to update grub configuration to pass the correct size as a kernel command line option or use stty something like: stty rows 62 cols 213 and it's resizes correctly. But is there any way to define the console's size outside the guest, in the VM's configuration? --- Thanks in advance, Roman Storozhenko

2 1

[libvirt-users] How to fit with Apparmor when upgrade to new libvirt version?
by Eddie Yen 24 May '17

24 May '17

Hi everyone, this is my first post at this mailing list. I have a question about upgrading libvirt, but also can fit to Apparmor. For example, I already installed KVM + libvirt from apt-get on Ubuntu 14.04. But the libvirt version is 1.2.9, so I want upgrade to 1.3.4 manually. Search the Internet, only few posts show how to edit so that can launch VM with Apparmor enabled. Most of posts says disable Apparmor is good idea, but I don't want to do that. I already done these things: 1. Replace apparmor profiles from examples/apparmor into /etc/apparmor.d 2. Use --with-apparmor and --with-apparmor-profiles to compile. But I still get virt-aa-helper DENIED message when I try to launch VM. Any ideas? Many thanks, Eddie.

1 0

[libvirt-users] Odd Virsh Capabilities response
by W Kern 24 May '17

24 May '17

Tried to migrate a few VMs from existing system to an older host that we threw in the rack to facilitate some server reshuffling. Migration complained about lack of capabilities (which I expected its an older server) but when I looked at what I should throw into the xml file I got a core2duo when it should have been at least Penryn or something (Xeon E5345) Of course the VMs work fine with that cpu def so its not a bit deal at the moment, but I am still curious Is it some sort of lookup table bug or is the server BIOS wonky? ----------------------------------------------------------- OS: Up to date U16 Xenial on an HP DL380 server. Stock libraries for that distro libvirt0/xenial-updates,now 1.3.1-1ubuntu10.8 amd64 [installed,automatic] qemu-kvm/xenial-updates,xenial-security,now 1:2.5+dfsg-5ubuntu10.14 amd64 [installed,automatic] QEMU Full virtualization cat /proc/cpu shows processor : 7 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Xeon(R) CPU E5345 @ 2.33GHz stepping : 11 microcode : 0xb4 cpu MHz : 2333.361 # virsh capabilities <capabilities> <host> <uuid>33393138-3335-5553-4538-31324e39414b</uuid> <cpu> <arch>x86_64</arch> <model>core2duo</model> <vendor>Intel</vendor> <topology sockets='2' cores='4' threads='1'/> <feature name='lahf_lm'/> <feature name='dca'/> <feature name='pdcm'/> <feature name='xtpr'/> <feature name='cx16'/> <feature name='tm2'/> <feature name='est'/> <feature name='vmx'/> <feature name='ds_cpl'/> <feature name='dtes64'/> <feature name='pbe'/> <feature name='tm'/> <feature name='ht'/> <feature name='ss'/> <feature name='acpi'/> <feature name='ds'/> <pages unit='KiB' size='4'/> <pages unit='KiB' size='2048'/> </cpu> <power_management> <suspend_mem/> <suspend_disk/> <suspend_hybrid/> </power_management> <migration_features> <live/> <uri_transports> <uri_transport>tcp</uri_transport> <uri_transport>rdma</uri_transport> </uri_transports> </migration_features> <topology> <cells num='1'> <cell id='0'> <memory unit='KiB'>32945368</memory> <pages unit='KiB' size='4'>8236342</pages> <pages unit='KiB' size='2048'>0</pages> <distances> <sibling id='0' value='10'/> </distances> <cpus num='8'> <cpu id='0' socket_id='0' core_id='0' siblings='0'/> <cpu id='1' socket_id='1' core_id='0' siblings='1'/> <cpu id='2' socket_id='0' core_id='2' siblings='2'/> <cpu id='3' socket_id='1' core_id='2' siblings='3'/> <cpu id='4' socket_id='0' core_id='1' siblings='4'/> <cpu id='5' socket_id='1' core_id='1' siblings='5'/> <cpu id='6' socket_id='0' core_id='3' siblings='6'/> <cpu id='7' socket_id='1' core_id='3' siblings='7'/> </cpus> </cell> </cells> </topology>

2 2

[libvirt-users] error message
by Eric Blake 24 May '17

24 May '17

[moderator note: re-posting to list with screenshot deleted, in order to meet list policy size constraints] hello, i am trying to install libvirt but getting some error . even though i am folowing your instructions and steps please help me out this

1 0

[libvirt-users] live migration
by Hossein Azizabadi 23 May '17

23 May '17

I’m interested with the modifications of live migration with all storage on KVM . I intend to ask, modification performed on which part of KVM? Do storage blocks migrate sequentialy? I need information about it

1 0

[libvirt-users] disabling dnsmasq
by Michael Ströder 22 May '17

22 May '17

HI! I have some virtual networks (not only bridged) in my libvirt 3.3.0 installation and therefore libvirt always starts dnsmasq. But I want to use my own DHCP and DNS servers which are correctly working for this VM infrastructure. How to completely disable dnsmasq in libvirt? Ciao, Michael.

2 1

[libvirt-users] Access to the wiki
by Jlozada2426 22 May '17

22 May '17

Hello, I was wondering if I could get access to the wiki. I want to contribute and help out. Jonathan Lozada De La Matta Linux System Administrator ask.sysadminonlinux.org Sent with [ProtonMail](https://protonmail.com) Secure Email.

2 1

[libvirt-users] <source mode='private'> for PPPoE?
by Michael Ströder 20 May '17

20 May '17

HI! I'm currently reviewing my (functional) setup with one physical ethernet interface eth1 of the host system being connected to a VDSL modem (not router!). A fully virtualized guest acts as router with this source mode: <source dev='eth1' mode='bridge'/> Reading the docs [1] I'm thinking whether mode='private' would be a better choice because the network packets are supposed to be sent exclusively to the VDSL modem and received from there. Any thoughts? Ciao, Michael. [1] https://libvirt.org/formatdomain.html#elementsNICSDirect

1 0

[libvirt-users] Duplicate reboot events
by Innus, Martins 18 May '17

18 May '17

Hi, Running on: $ cat /etc/redhat-release CentOS Linux release 7.3.1611 (Core) And: $ rpm -qa |grep libvirt libvirt-daemon-2.0.0-10.el7_3.5.x86_64 libvirt-daemon-driver-qemu-2.0.0-10.el7_3.5.x86_64 libvirt-daemon-driver-secret-2.0.0-10.el7_3.5.x86_64 libvirt-client-2.0.0-10.el7_3.5.x86_64 libvirt-daemon-driver-network-2.0.0-10.el7_3.5.x86_64 libvirt-daemon-driver-storage-2.0.0-10.el7_3.5.x86_64 libvirt-daemon-config-nwfilter-2.0.0-10.el7_3.5.x86_64 libvirt-daemon-config-network-2.0.0-10.el7_3.5.x86_64 libvirt-daemon-driver-nodedev-2.0.0-10.el7_3.5.x86_64 libvirt-2.0.0-10.el7_3.5.x86_64 libvirt-python-2.0.0-2.el7.x86_64 libvirt-daemon-driver-nwfilter-2.0.0-10.el7_3.5.x86_64 libvirt-daemon-driver-lxc-2.0.0-10.el7_3.5.x86_64 libvirt-daemon-driver-interface-2.0.0-10.el7_3.5.x86_64 I’m seeing duplicated reboot events when using the python event api. Use is simplified down to the simple test case attached. Running with that results in: $ sudo ./libvirt_events_single.py Reboot: Domain i-06945b37(21) 1494958504.72 Reboot: Domain i-06945b37(21) 1494958504.74 This is reproducible 100% of the time by doing /sbin/reboot from within a VM started with qemu-kvm. Any ideas on whats going on? Let me know if any other information would be useful. Thanks Martins

2 6

[libvirt-users] Callback re-registration after libvirtd restart
by Daniel Kučera 18 May '17

18 May '17

Hi all, I'm using libvirt-go and I following code to listen for lifecycle events: func event_listen() { log.Printf("event_listen %s", conf.Libvirt.LocalUrl) hv, err := libvirt.NewConnect(conf.Libvirt.LocalUrl) lifecycleCallback := func(c *libvirt.Connect, d *libvirt.Domain, event *libvirt.DomainEventLifecycle) { event_message(c, d, "lifecycle", event) } _, err = hv.DomainEventLifecycleRegister(nil, lifecycleCallback) if err != nil { log.Printf("unable to register event callback") return } log.Printf("Libvirt event listener started") go func() { for err == nil { err = libvirt.EventRunDefaultImpl() log.Printf("EventRunDefaultImpl err: %+v", err) } time.Sleep(time.Second) event_listen() }() } It works ok until I restart libvirtd (service libvirtd restart). After that, the inner go func waits some time and continues without error. But the callback is not working anymore. My question is, how can I detect hv reconnect (I guess it's happening in background) so I know when to reinitialize callbacks? Thanks. BR Daniel Kucera.

3 7

[libvirt-users] Poor network performance
by Sławomir Kapłoński 16 May '17

16 May '17

Hello, I have some problem with poor network performance on libvirt with qemu and openvswitch. I’m using libvirt 1.3.1, qemu 2.5 and openvswitch 2.6.0 on Ubuntu 16.04 currently. My connection diagram looks like below: +---------------------------+ +---------------------------+ | Net namespace | +------------------+ | OVS bridge | | | | | | | | | | VM | | | | | | | +----+---+ +----+-----+ +----+---+ | | +------+tap dev | | veth A +---------+ veth B | | | | +--------+ +----------+ +--------+ | | iperf -s<---------------------------------------------------------------------------+iperf -c | | | | | | | +------------------+ | | | | | | | | +---------------------------+ +---------------------------+ I haven’t got any QoS in tc configured on any interface. When I do this iperf test I have something about 150Mbps only. IMHO it should be something about 20-30 Gbps there. Other strange thing is that if I made more such VMS (each connected in same way, with own OVS bridge and own namespace) bandwidth is lower for each of them and it looks for me that summarize bandwidth is then something about 1Gbps (30 VMs - each got something about 30Mbps in such test). When I removed VM and added „tap dev” as internal port in OVS and made same test, then I had result about 30 Gbps. I have no idea what can be wrong there. Maybe someone of You had such problems earlier? One more thing, on different host with Ubuntu 14.04, OVS 2.0.2, Libvirt 1.3.1 and qemu 2.3 I don’t have this problem. Kernel on both hosts is 4.4.0 (from Ubuntu repo). Pozdrawiam / Best regards Sławek Kapłoński slawek(a)kaplonski.pl

2 3

[libvirt-users] strange error while trying to start ~60 domains
by Vasiliy Tolstov 09 May '17

09 May '17

Hi! I'm using latest libvirt release under debian jessie (kernel 4.4.x LTS) And after i'm try to start 65 domains i get 19 errors like: May 07 22:32:28 cn17 libvirtd[13836]: 2017-05-07 19:32:28.188+0000: 13839: error : virDBusCall:1570 : error from service: CreateMachine: Message did not receive a reply (timeout by message bus) I'm use qcow2 file as vm images on ssd disk with dir storage pool (local ssd disks). -- Vasiliy Tolstov, e-mail: v.tolstov(a)selfip.ru

2 2

Re: [libvirt-users] nwfilter and address of network ip address
by Dan 09 May '17

09 May '17

On Fri, May 5, 2017 at 4:29 PM, Nicolas Bock <nicolasbock(a)gmail.com> wrote: > Hi, > > I am running a webserver on the libvirt host and would like to add a > nwfilter such that a VM can access that server. The corresponding iptables > rule would look like this: > > iptables --append INPUT --in-interface virbr0 --destination 192.168.122.1 > --protocol tcp --dport 80 --jump ACCEPT > > where the network is using virbr0 and sits at 192.168.122.1. I don't want > to hardcode the host IP address in the nwfilter so that I can use that > filter for other networks. Is it possible to reference the host's IP > address in the filter? > > Thanks! > > Nick > > Hi Nick, I used to have similar question before too. Not sure if this could be helpful for you, Probably just use arp or arpscan. But to be more specific, if the domain name of the client (assuming you want to confine HTTP service to only a selected few clients, e.g., 192.168.122.1 in your case) is known, you probably could do $ virsh domiflist DOMAIN In particular, to get the IP address of a domain, it is something like the following: $ for MAC in `virsh domiflist <DOMAIN> | grep -o -E "([0-9a-f]{2}:){5}([0-9a-f]{2})"` ; do arp -e | grep $MAC | grep -o -P "^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}" ; done Alternatively, if the NETWORK name is known, IP addr can be obtained directly with $ virsh net-dhcp-leases NETWORK Another way to the get IP addr, if qemu guest agent is installed on the client, $ virsh domifaddr DOMAIN So that you can pass IP to the XML, and somehow you can trigger the update of NWFILTER with some magic, quite a hack. Or, you could just use client's MAC addr to define the filter: https://libvirt.org/formatnwfilter.html#nwfelemsRulesProtoMAC But I don't think there is direct way to specify a client's domain name in the NWFILTER XML definition as of now. P.S. I am new to libvirt, so everything above could be wrong. Dan _______________________________________________ > libvirt-users mailing list > libvirt-users(a)redhat.com > https://www.redhat.com/mailman/listinfo/libvirt-users >

3 6

[libvirt-users] nwfilter and address of network ip address
by Nicolas Bock 06 May '17

06 May '17

Hi, I am running a webserver on the libvirt host and would like to add a nwfilter such that a VM can access that server. The corresponding iptables rule would look like this: iptables --append INPUT --in-interface virbr0 --destination 192.168.122.1 --protocol tcp --dport 80 --jump ACCEPT where the network is using virbr0 and sits at 192.168.122.1. I don't want to hardcode the host IP address in the nwfilter so that I can use that filter for other networks. Is it possible to reference the host's IP address in the filter? Thanks! Nick

1 0

[libvirt-users] virsh snapshot delete / live-migration issues
by chihhang wu 05 May '17

05 May '17

Sorry, I made a mistake with my title. So I resend this email. ---------------------------------------------------------------------------------- Hi, I'd like to live-migration after snapshots deleted. But it doesn't work.Here's my steps: # virsh snapshot-create VM # virsh snapshot-delete VM Snapshot # virsh migrate --live VM qemu+ssh://localadmin@compute3/system And It comes out this message: # internal error: unable to execute QEMU command 'migrate-set-capabilities': There's a migration process in progress Even I create a no-metadata snapshot, it still have same error. After I reboot the VM, the live-migration finally works. Does it has any suggestion way or command to solve this problem without reboot the VM? Here's my environment: Ubuntu 14.04.5 LTS Compiled against library: libvirt 1.3.1 Using library: libvirt 1.3.1 Using API: QEMU 1.3.1 Running hypervisor: QEMU 2.5.0 Best regards, Chihhang Wu

2 1

[libvirt-users] Libvirtd freezes
by Stefano Ricci 04 May '17

04 May '17

Hello everyone I come back to ask for a hand to solve a problem that has affected me since October 2016 and I have not yet solved using libvirt. I thought I would solve it by going to a 4.9.x kernel with qemu 2.8.1.1 and with libvirt 3.2.0. Compile it all in a stable LFS environment version 7.9 and that all checks pass without errors. The strange thing is that the libvirtd process starts without errors but has arrived at the qemu process launch to understand the system's capabilities freezes until the following process is killed /usr/bin/qemu-system-x86_64 -S -none-user-config -nodefaults -nographic -machine none, accel = kvm: tcg -qmp unix: /var/lib/libvirt/qemu/capabilities.monitor.sock,server,nowait-pidfile/var/lib/libvirt/qemu/capabilities.Pidfile -daemonize, since that time libvirtd resumes running and can be used with virsh. Performing qemu independently of libvirt works regularly, creates and runs virtual machines smoothly. Thanks in advance Stefano Ricci

2 12

[libvirt-users] building virtual desktops with libvirt, KVM, SPICE and GNOME
by Daniel Pocock 02 May '17

02 May '17

Can anybody comment on how to host virtual desktops on a headless server using libvirt and KVM on the server and a SPICE client to access the virtual desktop? Is there a standard way of doing this? I've seen many fragments of information about how to do this but I didn't come across a single guide describing the entire solution. Search engines also return a lot of information about gaining remote access to a real physical desktop but that is not what I'm looking for. I've also come across many real-world scenarios where people are manually starting VNC server processes for each user on different ports but I was hoping to find out if there is a more standard way of doing this now. When I say "virtual desktop", the type of user experience I'm thinking about is that named users can run a SPICE client anywhere and always connect to the same host/desktop. E.g. if they leave some windows open, disconnect, go to another physical machine and reconnect with the same username they will see the same desktop with the same windows open. Regards, Daniel

2 2