- Users - libvirt List Archives

[libvirt-users] Advice request for building a centralized vm disk provider server
by Pasquale Dir 01 Apr '14

01 Apr '14

Hello again, sorry for mailing in so short time but I have another question about live migration. >From redhat documentation I see that the storage for a VM which has to be migrated has to be shared throught some mechanism and I have chosen iscsi. Anyway I now would need a way to dinamically create a new virtual machine and so a virtual hard disk, note that this vm will have to be migrated too. I though about this solution: I would share an HUGE iscsi target, this target is a logical hard drive containing many disk images which would be the VM's hard drives. I can easily clone a vm by making a simple file copy of a vm disk. This would allow me to create a VM by specifying a path like <iscsi target path>/disk1.img. I have two problems: 1)how to get an unique <iscsi target path>? On my phisical host it is mounted as a random uuid in the /media folder. 2)will migration work with a <file> instead of a <device> as hard disk field? If you have any better solution, even involing the change from iscsi to another technology, you are free to propose it to me. Sorry for bothering, Pasquale

1 0

[libvirt-users] Async migration: possible?
by Pasquale Dir 01 Apr '14

01 Apr '14

I am seeing that migrate() function is blocking, isn't there a non blocking function with a callback associated to it, so to report events like end migration, migration progress (job domain info informations) and so on? Pasquale

2 2

[libvirt-users] Ask for help about bug 1058149
by Qiang Guan 01 Apr '14

01 Apr '14

On Mar 31, 2014 at 21:23PM, Martin Kletzander wrote: > On Thu, Mar 27, 2014 at 10:07:43AM +0800, Qiang Guan wrote: >> Hi Martin, >> >> Is there any plan to fix the libvirt bug 1058149 which I reproduced on >> the libvirt master branch? >> > > Hi, > > Sorry for such late response, I'm currently busy working on whole > different part of the codebase. I'll try to see what I can do, but > you can always try asking on the list. Is there some version you were > not able to reproduce it on? > > Martin I've succeeded reproducing this problem on v1.1.4-maint ,v1.2.1-maint, v1.2.2 and the master branch. It seems that the problem always exsisted for libvirt under my testing scene. -- ------------ Jackie Best Regards

1 0

[libvirt-users] Programmatically force shutdown a guest: possible?
by Pasquale Dir 31 Mar '14

31 Mar '14

I am looking at the shutdown method, but if the guest system is a desktop system, like for example ubuntu, it just has the effect to show a box prompting the user for a shutdown/reboot/ and such. I could enter the guest and change this default behaviour and it actually works..but I'd like for a way to send a shutdown command without doing so. Is it possible?

3 4

[libvirt-users] Connecting libvirt to manually compiled QEMU
by Asadullah Hussain 31 Mar '14

31 Mar '14

Hello I have manually compiled a customized qemu (1.4.0) which runs fine on its own (create VM etc) but I want to access this qemu through libvirt (virt-manager, virsh etc). But the libvirt driver only looks into "/usr/bin" for qemu binaries, how can I tell libvirt to connect to my qemu which is placed at "/home/user/qemu" directory. -- Asadullah Hussain

3 6

[libvirt-users] Specify a disk image file which is on an iscsi target
by Pasquale Dir 31 Mar '14

31 Mar '14

Hello, for my vm I'd like to specify a disk image which would be on an iscsi target. Currently, by using targets directly, I'd just specify the path in /dev/disk/by-path/ip-192.168.1.2:3260-iscsi-iqn.myiqn but how to specify a file which is actually INSIDE this path, ex. /dev/disk/by-path/ip-192.168.1.2:3260-iscsi-iqn.myiqn/myVmImage.img ? I saw that there is another path in /media/myHomeUser/anId but do this id changes each time?

1 0

[libvirt-users] virsh attach-disk live guest problem
by john fisher 28 Mar '14

28 Mar '14

Hypervisor= ubuntu 12.04.2 guest = ubuntu 12.04.4 used virt-manager to create a new storage file in qcow2 format on hypervisor and then attach it to the guest. In virt-man I can see the disk is attached. trying virsh attach-disk I get a failure saying disk already attached virsh dump xml shows the disk at vdb as it should be but on the guest, fdisk -l shows no vdb and its not in /proc I found one reference to apparmor saying add /proc rw to hypervisor's apparmor libvirt conf But I see some release notes that might be about avoiding this ( sounds insecure) so... Do I have to stop and restart the guest, change apparmor or what? -- John Fisher

1 1

[libvirt-users] Using qemu+ssh on openSUSE 13.1/Tumbleweed
by Johannes Kastl 28 Mar '14

28 Mar '14

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi everyone, I am trying to get libvirt with qemu-kvm to work on my machines running openSUSE 13.1 / Tumbleweed. My question is in regard to using qemu+ssh, which would be my preference, as I already have a working ssh-key authentication with SSH-Agent. I set the permissions of the manage-socket to 0770, added my user to the libvirt group and set auth_unix_rw = "none" in /etc/libvirt/libvirtd.conf. I am now able to see all my VMs using virsh list --all (I had to set a default_uri in .config/libvirt/libvirt.conf to get any non-empty output). Trying from another machine (say target hostname is HOSTX, username testuser) with virsh -c qemu+ssh://testusers@HOSTX/system always results in >> error: failed to connect to the hypervisor error: internal error: >> received hangup / error event on socket I see this in /var/log/messages (lines will be wrapped in the mail, sorry): > 2014-03-24T21:25:55.023525+01:00 HOSTX sshd[3509]: Accepted > publickey for testuser from 192.168.79.8 port 54176 ssh2 > 2014-03-24T21:25:55.024744+01:00 HOSTX sshd[3509]: > pam_unix(sshd:session): session opened for user testuser by (uid=0) > 2014-03-24T21:25:55.025717+01:00 HOSTX kernel: [ 3162.225673] > type=1006 audit(1395692755.024:69): pid=3509 uid=0 old > auid=4294967295 new auid=1000 old ses=4294967295 new ses=24 res=1 > 2014-03-24T21:25:55.027501+01:00 HOSTX systemd[1]: Starting Session > 24 of user testuser. 2014-03-24T21:25:55.027918+01:00 HOSTX > systemd-logind[703]: New session 24 of user testuser. > 2014-03-24T21:25:55.028213+01:00 HOSTX systemd[1]: Started Session > 24 of user testuser. 2014-03-24T21:25:55.081719+01:00 HOSTX > sshd[3511]: Received disconnect from 192.168.79.8: 11: disconnected > by user 2014-03-24T21:25:55.082095+01:00 HOSTX sshd[3509]: > pam_unix(sshd:session): session closed for user testuser > 2014-03-24T21:25:55.086268+01:00 HOSTX systemd-logind[703]: Removed > session 24. I am curious about the 'Received disconnect from 192.168.79.8: 11: disconnected by user' thing. How can I fix this? Or how to debug this? How can I get more info from libvirt? /var/log/kvm is an empty directory, /var/log/libvirt/qemu just has logs for the different VMs. I have already tried to get polkit running, although I think it should not be used due to the auth_unix_rw = "none". Any help will be appreciated. Versions: openSUSE TUmbleweed (which is openSUSE 13.1 with some more recent software) qemu-kvm is version 1.7.90-19.1.x86_64 libvirt is 1.1.2 The following packes are installed: libvirt-1.1.2-2.18.3.x86_64 libvirt-client-1.1.2-2.18.3.x86_64 libvirt-daemon-1.1.2-2.18.3.x86_64 libvirt-daemon-config-network-1.1.2-2.18.3.x86_64 libvirt-daemon-config-nwfilter-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-interface-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-libxl-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-lxc-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-network-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-nodedev-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-nwfilter-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-qemu-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-secret-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-storage-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-uml-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-vbox-1.1.2-2.18.3.x86_64 libvirt-daemon-driver-xen-1.1.2-2.18.3.x86_64 libvirt-daemon-lxc-1.1.2-2.18.3.x86_64 libvirt-daemon-qemu-1.1.2-2.18.3.x86_64 libvirt-doc-1.1.2-2.18.3.x86_64 libvirt-glib-1_0-0-0.1.7-2.1.3.x86_64 libvirt-python-1.1.2-2.18.3.x86_64 Regards, Johannes - -- Have you ever noticed that the Klingons are all speaking unix? 'Grep ls awk chmod.'' 'Mknod ksh tar imap.' 'Wall fsck yacc!' (that last is obviously a curse of some sort). Gandalf Parker -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) Comment: Using GnuPG with SeaMonkey - http://www.enigmail.net/ iEYEARECAAYFAlM0CU8ACgkQzi3gQ/xETbKKyQCfVQ4XviDo5+92ivOIgx4lXIty 980An28BKHjfwZiRh5T2Yyj3wHwOLrYJ =ujio -----END PGP SIGNATURE-----

1 2

[libvirt-users] host crashes "unable to handle paging request"
by Raphael Bauduin 28 Mar '14

28 Mar '14

Hi, we have regular crashed of a kvm host with the error "unable to handle paging request". Can this be due to memory over-commitment even if some memory is still used by the kernel for caches and buffers? (collectd graph shows no free memory, with 15G used, very little buffers, and 1G cache). There are 32GB of swap, of which only 150MB are used. I suspect might be the direction to search to find the cause, but would be happy to learn from people versed in the kernel behaviour to confirm or reject my hypothesis. Below is the full error. Thanks! Raph 745 Mar 23 14:27:37 sMaster01 kernel: [241450.355339] BUG: unable to handle kernel paging request at ffff8804c001fade 746 Mar 23 14:27:37 sMaster01 kernel: [241450.355384] IP: [<ffffffff8117e9e9>] bio_check_eod+0x29/0xcd 747 Mar 23 14:27:37 sMaster01 kernel: [241450.355433] PGD 1002063 PUD 0 748 Mar 23 14:27:37 sMaster01 kernel: [241450.355464] Oops: 0000 [#1] SMP 749 Mar 23 14:27:37 sMaster01 kernel: [241450.355496] last sysfs file: /sys/devices/system/cpu/cpu15/ topology/thread_siblings 750 Mar 23 14:27:37 sMaster01 kernel: [241450.355551] CPU 4 751 Mar 23 14:27:37 sMaster01 kernel: [241450.355577] Modules linked in: ipt_MASQUERADE iptable_nat nf_nat nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack ipt_REJECT xt_tcpudp kvm_amd kvm ip6table_filter ip6_tables iptable_fi lter ip_tables x_tables tun nfsd exportfs nfs lockd fscache nfs_acl auth_rpcgss sunrpc bridge stp bonding dm_round_robin dm_multipath scsi_dh loop snd_pcm snd_timer snd soundcore snd_page_alloc serio_raw evdev tpm_tis tpm tpm_bios p smouse pcspkr amd64_edac_mod edac_core button edac_mce_amd shpchp i2c_piix4 container pci_hotplug i2c_core processor ext3 jbd mbcache dm_mirror dm_region_hash dm_log dm_snapshot dm_mod sd_mod crc_t10dif mptsas mptscsih mptbase lpfc ehci_hcd scsi_transport_fc tg3 scsi_tgt scsi_transport_sas ohci_hcd libphy scsi_mod usbcore nls_base thermal fan thermal_sys [last unloaded: scsi_wait_scan] 752 Mar 23 14:27:37 sMaster01 kernel: [241450.356084] Pid: 3557, comm: kjournald Not tainted 2.6.32.61vanilla #1 PRIMERGY BX630 S2 753 Mar 23 14:27:37 sMaster01 kernel: [241450.356141] RIP: 0010:[<ffffffff8117e9e9>] [<ffffffff8117e9e9>] bio_check_eod+0x29/0xcd 754 Mar 23 14:27:37 sMaster01 kernel: [241450.356196] RSP: 0018:ffff8804229abba0 EFLAGS: 00010202 755 Mar 23 14:27:37 sMaster01 kernel: [241450.356228] RAX: ffff8804c001fad6 RBX: ffff8802e7235080 RCX: 00011200061e5110 756 Mar 23 14:27:37 sMaster01 kernel: [241450.356279] RDX: 0000000000000008 RSI: 0000000000000008 RDI: ffff8802e7235080 757 Mar 23 14:27:37 sMaster01 kernel: [241450.356331] RBP: ffff8802e7235080 R08: 0000000000000000 R09: ffff880425c54c00 758 Mar 23 14:27:37 sMaster01 kernel: [241450.356383] R10: 0000000000000003 R11: 00000000022e539e R12: ffff8802e7235080 759 Mar 23 14:27:37 sMaster01 kernel: [241450.356434] R13: ffff8802e7235080 R14: ffff880425c54c00 R15: ffff8802e6281850 760 Mar 23 14:27:37 sMaster01 kernel: [241450.356486] FS: 00007faa6a757820(0000) GS:ffff88000fc80000(0000) knlGS:0000000000000000 761 Mar 23 14:27:37 sMaster01 kernel: [241450.356540] CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b 762 Mar 23 14:27:37 sMaster01 kernel: [241450.356573] CR2: ffff8804c001fade CR3: 00000000cc11f000 CR4: 00000000000006e0 763 Mar 23 14:27:37 sMaster01 kernel: [241450.356628] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 764 Mar 23 14:27:37 sMaster01 kernel: [241450.356681] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 765 Mar 23 14:27:37 sMaster01 kernel: [241450.356733] Process kjournald (pid: 3557, threadinfo ffff8804229aa000, task ffff88041490a300) 766 Mar 23 14:27:37 sMaster01 kernel: [241450.356788] Stack: 767 Mar 23 14:27:37 sMaster01 kernel: [241450.356812] ffff880415382c00 0000000100000285 ffff8804229abfd8 0000000000005186 768 Mar 23 14:27:37 sMaster01 kernel: [241450.356852] <0> 0000000000000000 000000000f1c2776 ffff8804128efa38 ffff8802e7235080 769 Mar 23 14:27:37 sMaster01 kernel: [241450.356913] <0> ffff8802e7235080 ffff8802e7235080 ffff8800cdacae40 ffffffff8117eb5a 770 Mar 23 14:27:37 sMaster01 kernel: [241450.356993] Call Trace: 771 Mar 23 14:27:37 sMaster01 kernel: [241450.357021] [<ffffffff8117eb5a>] ? generic_make_request+0xcd/0x2f9 772 Mar 23 14:27:37 sMaster01 kernel: [241450.357058] [<ffffffff810b6034>] ? mempool_alloc+0x55/0x106 773 Mar 23 14:27:37 sMaster01 kernel: [241450.357091] [<ffffffff8117ee5c>] ? submit_bio+0xd6/0xf2 774 Mar 23 14:27:37 sMaster01 kernel: [241450.357125] [<ffffffff8110d83f>] ? submit_bh+0xf5/0x115 775 Mar 23 14:27:37 sMaster01 kernel: [241450.357158] [<ffffffff8110edc0>] ? sync_dirty_buffer+0x51/0x93 776 Mar 23 14:27:37 sMaster01 kernel: [241450.357196] [<ffffffffa01727c7>] ? journal_commit_transaction+0xaa6/0xe4f [jbd] 777 Mar 23 14:27:37 sMaster01 kernel: [241450.357252] [<ffffffffa0175194>] ? kjournald+0xdf/0x226 [jbd] 778 Mar 23 14:27:37 sMaster01 kernel: [241450.357288] [<ffffffff810651de>] ? autoremove_wake_function+0x0/0x2e 779 Mar 23 14:27:37 sMaster01 kernel: [241450.357324] [<ffffffffa01750b5>] ? kjournald+0x0/0x226 [jbd] 780 Mar 23 14:27:37 sMaster01 kernel: [241450.357357] [<ffffffff81064f11>] ? kthread+0x79/0x81 781 Mar 23 14:27:37 sMaster01 kernel: [241450.357391] [<ffffffff81011baa>] ? child_rip+0xa/0x20 782 Mar 23 14:27:37 sMaster01 kernel: [241450.357425] [<ffffffff81016568>] ? read_tsc+0xa/0x20 783 Mar 23 14:27:37 sMaster01 kernel: [241450.357456] [<ffffffff81064e98>] ? kthread+0x0/0x81 784 Mar 23 14:27:37 sMaster01 kernel: [241450.357487] [<ffffffff81011ba0>] ? child_rip+0x0/0x20 785 Mar 23 14:27:37 sMaster01 kernel: [241450.357517] Code: 5c c3 41 55 49 89 fd 41 54 55 53 48 83 ec 38 65 48 8b 04 25 28 00 00 00 48 89 44 24 28 31 c0 85 f6 0f 84 86 00 00 00 48 8b 47 10 <48> 8b 40 08 48 8b 40 68 48 c1 f8 09 74 74 89 f2 48 8b 0f 48 39 786 Mar 23 14:27:37 sMaster01 kernel: [241450.357738] RIP [<ffffffff8117e9e9>] bio_check_eod+0x29/0xcd 787 Mar 23 14:27:37 sMaster01 kernel: [241450.357772] RSP <ffff8804229abba0> 788 Mar 23 14:27:37 sMaster01 kernel: [241450.357799] CR2: ffff8804c001fade 789 Mar 23 14:27:37 sMaster01 kernel: [241450.358183] ---[ end trace 608fcf1f5a482549 ]--- -- Web database: http://www.myowndb.com Free Software Developers Meeting: http://www.fosdem.org

1 1

[libvirt-users] LXC + passthrough mount and host filesystem-cache
by James R. Leu 28 Mar '14

28 Mar '14

Hello, I'm using libvirt to build/run LXC instances. My LXC instances use passthrough filesystem mounts. When I try to do large file systems operations (ex tar or rsync) the file systems cache on the host spikes and causes the OOM handler to run and kills processes in the LXC. Has anyone else seen this? Is there a way around this? At this point I'm resorting to running a cron job that dumps the filesystem cache every 5 minutes. The result is the filesystem cache on the host never grows too large and OOM never runs against LXC processes. The obvious down fall is that I'm killing my filesystem performance by duming the cache. I'm currently running libvirt 1.0.0 -- James R. Leu

2 2