- Users - libvirt List Archives

[libvirt-users] libvirt/dnsmasq is not adhering to static DHCP assignments
by Christian Kujau 09 Aug '19

09 Aug '19

This is basically a continuation of an older posting[0] I found, but apparently no solution has been posted. So, I'm trying to setup static DHCP leases with dnsmasq that is being started by libvirtd: ----------------------------------------------------------- $ sudo virsh net-dumpxml --network default <network> [...] <dns enable='no'/> <ip address='192.168.122.1' netmask='255.255.255.0'> <dhcp> <range start='192.168.122.130' end='192.168.122.250'/> <host mac='08:00:27:e2:81:39' name='f30' ip='192.168.56.139'/> ----------------------------------------------------------- And the domain is indeed being started with that hardware address: ----------------------------------------------------------- $ virsh dumpxml --domain f30 | grep -B1 -A3 mac\ a <interface type='bridge'> <mac address='08:00:27:e2:81:39'/> <source bridge='virbr0'/> <target dev='tap0'/> <model type='virtio'/> ----------------------------------------------------------- But for some reason the domain gets a different address assigned, albeit from the correct DHCP pool: ----------------------------------------------------------- $ ssh 192.168.122.233 "ip addr show" [...] 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000 link/ether 08:00:27:e2:81:39 brd ff:ff:ff:ff:ff:ff inet 192.168.122.233/24 brd 192.168.122.255 scope global dynamic enp1s0 ----------------------------------------------------------- See below for even more details. I've removed the virbr0.status file, restarted the network, still it gets this .233 address instead of the configured .139. Also, I'm not able to add a debug log to the libvirtd/dnsmasq instance, although it should be possible[1], the xmlns:dnsmasq stanza is overwritten as soon as the configuration is saved. Short of a debug (queries) log, strace on the libvirtd/dnsmasq process reveals that the client seems to request this strange IP address: 23002 19:01:59.053558 write(11, "<30>Jul 31 19:01:59 dnsmasq-dhcp[23002]: DHCPREQUEST(virbr0) 192.168.122.233 08:00:27:e2:81:39 ", 95) = 95 23002 19:01:59.054024 write(11, "<30>Jul 31 19:01:59 dnsmasq-dhcp[23002]: DHCPACK(virbr0) 192.168.122.233 08:00:27:e2:81:39 f30", 94) = 94 So I restarted the dhclient process in the domain (a freshly installed Fedora 30), removed all state files with "192.168.122.233" in them, but still the domains gets assigned the .233 instead of the .139 address. The next step would be to disable libvirtd/dnsmasq altogether and run my own dnsmasq instance, but I wanted to avoid that. Any idead on where to look next? Thanks, Christian. [0] https://www.redhat.com/archives/libvirt-users/2017-October/msg00070.html [1] https://libvirt.org/formatnetwork.html#elementsNamespaces # cd /var/lib/libvirt/dnsmasq/ # grep -r . . | fgrep -v \# ./default.conf:strict-order ./default.conf:port=0 ./default.conf:pid-file=/var/run/libvirt/network/default.pid ./default.conf:except-interface=lo ./default.conf:bind-dynamic ./default.conf:interface=virbr0 ./default.conf:dhcp-range=192.168.122.130,192.168.122.250,255.255.255.0 ./default.conf:dhcp-no-override ./default.conf:dhcp-authoritative ./default.conf:dhcp-lease-max=121 ./default.conf:dhcp-hostsfile=/var/lib/libvirt/dnsmasq/default.hostsfile ./default.hostsfile:08:00:27:e2:81:39,192.168.56.139,f30 ./virbr0.status:[ ./virbr0.status: { ./virbr0.status: "ip-address": "192.168.122.233", ./virbr0.status: "mac-address": "08:00:27:e2:81:39", ./virbr0.status: "hostname": "f30", ./virbr0.status: "client-id": "ff:27:e2:81:39:00:04:8c:ad:c4:7d:04:e8:4b:de:93:4b:76:d8:75:82:86:c8", ./virbr0.status: "expiry-time": 1564628519 ./virbr0.status: } ./virbr0.status:] -- BOFH excuse #40: not enough memory, go get system upgrade

2 2

[libvirt-users] Reg: content of disk is not reflecting in host.
by bharath paulraj 07 Aug '19

07 Aug '19

Hi Team, I am doing a small testing and I don't know if my expectation is correct or not. Pardon me if I am ignorant. I created a VM and the VM is running. In the hypervisor I have created ".img" file and attached this .img file to the VM. My expectation is that, if VM is writing files to the attached disk, then it should reflect in the .img file created in the hypervisor. But It is not working as my expectation. Please correct me if my expectation is wrong. Steps: 1. Created disk.img in the hypervisor using the command: dd if=/dev/zero of=disk.img bs=1M count=50; mkfs ext3 -F disk.img 2. Attached the disk to the running VM using the command: virsh attach-disk <Domain-Name> --source disk.img --target vdb --live 3. In the VM, I mounted the disk and created few files. 4. In the hypervisor, I mounted the disk.img to check if the file created in the VM exists in the .img file. >> I am not able to see those files. Regards, Bharath

2 2

[libvirt-users] Dev wanted regarding Ceph and/or libxl related bugs and feature requests
by Julian Sternberg 03 Aug '19

03 Aug '19

Hi there, anybody with ceph api and/or libxl skills able to implement following feature and fix bug requests: https://bugzilla.redhat.com/show_bug.cgi?id=1729408 https://bugzilla.redhat.com/show_bug.cgi?id=1731159 https://bugzilla.redhat.com/show_bug.cgi?id=1732306

1 0

[libvirt-users] Detach disk from VM - virsh (working) vs. PHP (not working)
by Jan Horak 03 Aug '19

03 Aug '19

Hi all, i created a script in PHP for create a virtual server with two QCOW2 discs … one is our system for installation and second is target system. After successfully instalation (create a blank Debian system, prepare all files and grub partitions) i need a restart virtual without a installation disk. If i use Virsh: detach-disk --domain debian-test2 --persistent --target vda reset debian-test2 everything works well. If i use a PHP, there is „complicated“ way and „simple“ way: 1, complicated: libvirt_domain_destroy($domain); libvirt_domain_undefine($domain); $xml = domain_create_xml($name,$uuid,$memory,$cpu,$vnc,$mac); $domain = libvirt_domain_define_xml($server->conn, $xml); libvirt_domain_disk_add($domain, "/users/".$name.".img", "vdb", "virtio", "qcow2", NULL); libvirt_domain_create($domain); (or instead libvirt_domain_disk_add i can define disk directly in XML) But in this case, the server will not boot (GRUB error) 2, simple: libvirt_domain_disk_remove($domain,“vda“); libvirt_domain_reboot($domain); The problem of this solution is thats not working. The remove disk is failing with error „Unable attach disk“ - i looks to source code, and yes, there is a mystake with „attach“/„detach“, but main problem i see in log from libvirt: Aug 1 02:57:05 ry libvirtd[19051]: missing source information for device vda I try to put source detail to xml in source of PHP module libvirt-domain.c: 822 if (asprintf(&newXml, 823 " <disk type='file' device='disk'>\n" 824 " <target dev='%s'/>\n" 825 " </disk>", dev) < 0) { 826 set_error("Out of memory" TSRMLS_CC); 827 goto error; 828 } but my attempts was unsuccesfull (i’m not C programmer). Questions: A, why complicated way is not working and system dont want boot (GRUB error) if virsh works fine B, why libvirt_domain_disk_remove is not work ? I use libvirt and libvirt-php latest from git. Thank you, Jan

4 5

[libvirt-users] Researching why different cache modes result in 'some' guest filesystem corruption..
by vincent＠cojot.name 02 Aug '19

02 Aug '19

Hi All, I've been chasing down an issue in recent weeks (my own lab, so no prod here) and I'm reaching out in case someone might have some guidance to share. I'm running fairly large VMs (RHOSP underclouds - 8vcpu, 32gb ram, about 200gb single disk as a growable qcow2) on some RHEL7.6 hypervisors (kernel 3.10.0-927.2x.y, libvirt 4.5.0, qemu-kvm-1.5.3) on top of SSD/NVMe drives with various filesystems (vxfs, zfs, etc..) and using ECC RAM. The issue can be described as follows: - the guest VMs work fine for a while (days, weeks) but after a kernel update (z-stream) comes in, I am often greeted by the following message immediately after rebooting (or attempting to reboot into the new kernel): "error: not a correct xfs inode" - booting the previous kernel works fine and re-generating the initramfs for the new kernel (from the n-1 kernel) does not solve anything. - if booted from an ISO, xfs_repair does not find errors. - on ext4, there seems to be some kind of corruption there too. I'm building the initial guest image qcow2 for those guest VMs this way: 1) start with a rhel-guest image (currently rhel-server-7.6-update-5-x86_64-kvm.qcow2) 2) convert to LVM by doing this: qemu-img create -f qcow2 -o preallocation=metadata,cluster_size=1048576,lazy_refcounts=off final_guest.qcow2 512G virt-format -a final_guest.qcow2 --partition=mbr --lvm=/dev/rootdg/lv_root --filesystem=xfs guestfish --ro -a rhel_guest.qcow2 -m /dev/sda1 -- tar-out / - | \ guestfish --rw -a final_guest.qcow2 -m /dev/rootdg/lv_root -- tar-in - / 3) use "final_guest.qcow2" as the basis for my guests with LVM. After chasing down this issue some more and attempting various things (build the image on Fedora29, build a real XFS filesystem inside a VM and use the generated qcow2 as a basis instead of virt-format).. ..I've noticed that the SATA disk of each of those guests were using 'directsync' (instead of 'Hypervisor Default'). As soon as I switched to 'None', the XFS issues disappeared and I've now applied several consecutive kernel updates without issues. Also, 'directsync' or 'writethrough', while providing decent performance, both exhibited the XFS 'corruption' behaviour.. Only 'none' seem to have solved that. I've read the docs but I thought it was OK to use those modes (UPS, Battery-Backed RAID, etc..) Does anyone have any idea what's going on or what I may be doing wrong? Thanks for reading, Vincent

2 1

[libvirt-users] Why librbd disallow VM live migration if the disk cache mode is not none or directsync
by Ming-Hung Tsai 01 Aug '19

01 Aug '19

I'm curious that why librbd sets this limitation? The rule first appeared in librbd.git commit d57485f73ab. Theoretically, a write-through cache is also safe for VM migration, if the cache implementation guarantees that cache invalidation and disk write are synchronous operations. For example, I'm using Ceph RBD images as VM storage backend. The Ceph librbd supports synchronous write-through cache, by setting rbd_cache_max_dirty to zero, and setting rbd_cache_block_writes_upfront to true, thus it would be safe for VM migration. Is that true? Any suggestion would be appreciated. Thanks. Ming-Hung Tsai

2 3

[libvirt-users] Libvirt Virtualization Problem
by 王金磊 30 Jul '19

30 Jul '19

Dear Sir/Madam: I use virt-manager to create a new virtual machine(it's name is generic), then I edit generic to set it's cpu_mode as 'host-model', and I start it, but when dumpxml generic, it's cpu_mode is changed as 'custom'. I want to know that why it is? And what's the principle of dumpxml? Thanks! A user of libvirt

2 1

[libvirt-users] virsh command to list CPU,Memory and Storage
by Kaushal Shriyan 30 Jul '19

30 Jul '19

Hi, Is there a way to find out total CPU's, Memory and Storage using virsh command? For example virsh list --all list out all VM's 11 dockerregistry01 running 12 gitlab running Any help will be highly appreciated. Thanks in Advance. Best Regards, Kaushal

3 2

[libvirt-users] <VM LIVE Migration> <Sync conntrack entries>
by bharath paulraj 26 Jul '19

26 Jul '19

Hi Team, I am using QEMU/KVM for launching VMs and libvirt to govern those VMs. I would like to synchronise the connection tracking entries specific to the VM during the VM LIVE migrations. It is required when the firewall is implemented at the host level like libvirt's "network filters". If stateful firewall is enabled, then unless these connection tracking entries are synchronised, all the connections to the VM are lost and all TCP connections should be reestablished. Is there any option already available? I don't think current libvirt hooks are helpful, as VM pause in the source hypervisor and VM on in the destination hypervisor is done by QEMU and it does not wait for any application that needs to sync-up some metadata — In my case, it is conntrack entries. Also I tried with the existing hooks - stop, release, startcpus and nothing worked well. Has anybody came across similar scenario? If yes, how you overcome this? -- Regards, Bharath

2 1

[libvirt-users] Define libvirt network portgroup with native untagged.
by Tanmoy Sinha 26 Jul '19

26 Jul '19

Hi, I have created a libvirt network on top of a OVS bridge (named vlan-br) which receives all VLAN tagged packets, i.e. connected to a trunk port. The definition xml is below. What I want to achieve in the portgroup definition 'trunk-native-1221' is to allow 1221 as untagged/native but rest all VLAN as tagged. The following portgroup definition works, but I don't want to enumerate all the tagged VLANs in the portgroup definition. I understand what libvirt does on the underlying OVS bridge, once a guest interface (say vNetX) is attached to the portgroup is to set the vnetX with tag=1221 and vlan_mode= native-untagged and it sets trunk = [1222,1221,1223,1224]. Now if I go and clear the trunk setting on the OVS bridge for that interface I am able to see both tagged and untagged (1221) packets on the guest. This is exactly what I want to achieve in the libvirt network definition, i.e. have one untagged VLAN and* allow all other VLANs without having to enumerate them in the portgroup definition, as that is hard to maintain. * <network> <name>kvm-core-net</name> <bridge name = 'vlan-br'/> <forward mode = 'bridge'/> <virtualport type='openvswitch'/> ....... <portgroup name='*trunk-native-1221*'> <vlan trunk='yes'> <tag id='1222'/> <tag id='1221' nativeMode='untagged'/> <tag id='1223'/> <tag id='1224'/> </vlan> </portgtroup> ..... </network> Regards Tanmoy Sinha

1 0