August 2024 - Users - oVirt List Archives

logical names of the disks are not created sequentially with ansible ovirt.ovirt.ovirt_disk
by marek 29 Aug '24

29 Aug '24

Hello, I am creating VM, and disks on VM, and I am occurring a problem, where the logical names of the disks are probably not created / assigned sequentially. This is the outcome that I would expect: disk_name | logical_name -------------------------------- os_disk | /dev/sda vm.example.com_4 | /dev/sdb vm.example.com_3 | /dev/sdc vm.example.com_2 | /dev/sdd vm.example.com_1 | /dev/sde However I have experienced on multiple occasions that the logical names are mixed, for example like this (see row 1 and 2): disk_name | logical_name -------------------------------- os_disk | /dev/sdb vm.example.com_4 | /dev/sda vm.example.com_3 | /dev/sdc vm.example.com_2 | /dev/sdd vm.example.com_1 | /dev/sde I have tried to add state: "present" and wait: true parameters for each disk, but this has not helped, and I am still facing the same problem. ansible snippet below any tips/advices where to dig? Steps to reproduce: --- - name: Auth ovirt ovirt.ovirt.ovirt_auth: url: https://ovirt.example.com/ovirt-engine/api # ... - name: Create vm vm.example.com ovirt.ovirt.ovirt_vm: high_availability: true auth: "{{ ovirt_auth }}" state: stopped cpu_cores: "2" memory: "6GiB" memory_max: "6GiB" name: "vm.example.com" template: "linux" cluster: "prod" type: server graphical_console: headless_mode: true operating_system: "linux" nics: - name: eth0 profile_name: "vlan" cloud_init_persist: true cloud_init: dns_servers: "1.1.1.1 1.0.0.1" host_name: "vm.example.com" user_name: root root_password: "ROOT_PASSWORD" authorized_ssh_keys: "SSH_KEY" cloud_init_nics: - nic_name: enp1s0 nic_boot_protocol: static nic_ip_address: "10.0.0.1" nic_netmask: "255.255.255.0" nic_gateway: "10.0.0.254" - ovirt.ovirt.ovirt_disk: name: os_disk bootable: true vm_name: "vm.example.com" auth: "{{ ovirt_auth }}" state: "present" wait: true - ovirt.ovirt.ovirt_disk: name: "vm.example.com_4" auth: "{{ ovirt_auth }}" vm_name: "vm.example.com" size: "2GiB" format: cow interface: virtio_scsi storage_domain: "STORAGE_DOMAIN" state: "present" wait: true - ovirt.ovirt.ovirt_disk: name: "vm.example.com_3" auth: "{{ ovirt_auth }}" vm_name: "vm.example.com" size: "5GiB" format: cow interface: virtio_scsi storage_domain: "STORAGE_DOMAIN" state: "present" wait: true - ovirt.ovirt.ovirt_disk: name: "vm.example.com_2" auth: "{{ ovirt_auth }}" vm_name: "vm.example.com" size: "20GiB" format: cow interface: virtio_scsi storage_domain: "STORAGE_DOMAIN" state: "present" wait: true - ovirt.ovirt.ovirt_disk: name: "vm.example.com_1" auth: "{{ ovirt_auth }}" vm_name: "vm.example.com" size: "5GiB" format: cow interface: virtio_scsi storage_domain: "STORAGE_DOMAIN" state: "present" wait: true - ovirt.ovirt.ovirt_vm: auth: "{{ ovirt_auth }}" state: running name: "vm.example.com"

1 1

Re: VM frozen during snapshot removing
by Domen Cesnik 28 Aug '24

28 Aug '24

Hi, This looks like a bug. We had a same issue with few VMs After Veeam support was involved they redirected us to Oracle as we use OLVM. There was in logs 2024-08-17 01:06:12.1818 00022 [4184461] WARN | [RhvBackupManager]: Failed to create incremental backup, fallback to full backup: RhevClient.Exceptions.RhevServerException: Cannot backup VM: Disk is locked. Please try again later. THe same error was with Vinchin.

1 0

Re: VM frozen during snapshot removing
by Domen Česnik 28 Aug '24

28 Aug '24

Hi, This looks like a bug as we have a same problem. We involved the Veeam support and they redirected us to Oracle as we use oracle version. This is happening on migrated v2v servers with centos7 [https://www.humanfrog.com/img/logo_hand.svg] Domen Česnik domen.cesnik(a)humanfrog.com<mailto:domen.cesnik@humanfrog.com> 041 550 772<tel:041%20550%20772> Humanfrog d.o.o. Dunajska 5, 1000 Ljubljana, Slovenia www.humanfrog.com<https://www.humanfrog.com>

1 0

oVirt networks
by Enrico Becchetti 28 Aug '24

28 Aug '24

Dear all, Ineed your help to understand how to configure the network of a new oVirt cluster. Mynew system will have a 4.3 engine thatruns in a virtual machine, andsome Dell R7525 AMD EPYC hypervisors, eachholding two 4-port PCI network cards. These servers will have node-ovirt image again in version 4.3. As for the network, there are two HPE Aruba 2540G, non-stackable, with 24 1Gbs ports and 2 10Gbs uplinks to the star center. This is a simplified scheme: My goal is to make the most of the server's 8 ethernet interfaces to have both reliability and maximum possible throughput. This cluster will have two virtual networks, one forovirt management and one for the traffic of individual virtual machines. With that said here's what my idea is. I would like to have two links aggregated by 4Gbs, one for ovrtmgt and the other for vmnet. With the ovirt web interface I can createan active-passive "Mode 1" bond, but this won'tallow me to go beyond 1Gbs. Alternatively I could create a "Mode 4" bond 802.3ad but unfortunately the switches are not stacked and therefore not even this solution applies. This is an example with active passive configuration: Can you tell me if ovirt can generate//nested bonds? Or do you have other solutions ? Thanks a lot ! Best Regards Enrico -- _______________________________________________________________________ Enrico Becchetti Servizio di Calcolo e Reti Istituto Nazionale di Fisica Nucleare - Sezione di Perugia Via Pascoli,c/o Dipartimento di Fisica 06123 Perugia (ITALY) Phone:+39 075 5852777 Skype:enrico_becchetti Mail: Enrico.Becchetti<at>pg.infn.it Pagina web personale: https://www.pg.infn.it/home/enrico-becchetti/ ______________________________________________________________________

5 6

boot from cdrom & error code 0005
by edp＠maddalena.it 21 Aug '24

21 Aug '24

Hi. I have created a new storage domain (data domain, storage type nfs) to use it to upload iso images. I have so uploaded a new iso and then attach the iso to a new vm. But when I try to boot the vm I obtain this error: booting from dvd/cd... boot failed: could not read from cdrom (code 0005) no bootable device The iso file has been uploaded with success in the data storage domain and so the vm lets my attach the iso to the vm in the boot settings. Can you help me? Thank you

4 6

Listing all roles/permissions of a user over REST
by sameer.save＠rackwareinc.com 21 Aug '24

21 Aug '24

Is there a way we could list all the permission a user has once its authenticated? i am clueless about if we can check wether a particular user can create a disk/vm/network profile.

1 0

Re: How to switch oVirt from Cockpit from RHEL9?
by Ruber van der Linden 20 Aug '24

20 Aug '24

Cockpit is a control panel for server management. The virtualization is only a part of it. You can have both Cockpit and oVirt installed on the same server, as they serve for different purposes. If you already use KVM and have any VM running, installing oVirt over it may cause some trouble. I suggest starting with a clean server using the self hosted engine.

1 0

Best way to shutdown ovirt-engine.
by 1990w＠naver.com 17 Aug '24

17 Aug '24

Hi all, Ovirt-engine was configured as vm. There is a problem because Ovirt-engine's resources are too small. So I'm going to add resources after I shutdown Ovirt-engine. What is the best way to shut down Ovirt-engine? Is it okay to stop engine service and shutdown vm?

2 1

VM frozen during snapshot removing
by Kalil de A. Carvalho 14 Aug '24

14 Aug '24

Hello all. In my company we have a oVirt 4.5.4-1.el9 running and a Vinchin Backup & Recovery 8.0 as a backup solution. A strange situation. Same VM's stopping working, frozen, when the snaphost it is been removing. And when this happened we must kill the process directly in the host because we can't do this on oVirt web management. I just found some information on message log, notting on vdsm or libvirt logs. Here what is in the message log: Aug 13 21:49:26 HOST libvirtd[5304]: Cannot start job (modify, none, none) in API remoteDispatchDomainSetSchedulerParameters for domain VM; current job is (query, none, none) owned by (5332 remoteDispatchConnectGetAllDomainStats, 0 <null>, 0 <null> (flags=0x0)) for (45s, 0s, 0s) Aug 13 21:49:56 HOST libvirtd[5304]: Cannot start job (modify, none, none) in API remoteDispatchDomainSetSchedulerParameters for domain VM; current job is (query, none, none) owned by (5332 remoteDispatchConnectGetAllDomainStats, 0 <null>, 0 <null> (flags=0x0)) for (75s, 0s, 0s) Has anyone passed for this situation? Bast regards. -- Atenciosamente, Kalil de A. Carvalho

1 0

moving VMs from old oVirt to new
by Diggy Mc 14 Aug '24

14 Aug '24

Since export domains have been deprecated starting with oVirt v4.4, what is the preferred/best method for moving VMs from an old oVirt environment into a new oVirt environment? The two environments are totally independent of each other and therefore need an intermediary.

4 5