- Users - oVirt List Archives

Re: [External] : Custom OMVF for VM
by Marcos Sungaila 11 Jul '25

11 Jul '25

Hi Domien, Possible it is, but you will need to test this deployment. The firmware json files are stored under /usr/share/qemu/firmware on each KVM host, and the fd files are stored under /usr/share/edk2/ovmf/ on the KVM hosts, too. As far as I understand, and you need to confirm it, the Chipset/Firmware Type option in the web UI is retrieved from the PostgreSQL database requiring you to add additional entries to it so the web portal will display the corresponding options to deploy this appliance. The good point, by going this path you will be able to create a great blog article describing how to create a custom entry in oVirt to deploy Sonicwall in it. Marcos From: Domien Van Rompaey via Users <users(a)ovirt.org> Sent: Tuesday, July 8, 2025 10:45 AM To: users(a)ovirt.org Subject: [External] : [ovirt-users] Custom OMVF for VM Hi We want to use the Sonicwall NSv appliance on our oVirt Cluster. Since version 7.1.1 they require the installation of a custom OMVF (UEFI Firmware). Sonicwall KB: https://www.sonicwall.com/support/knowledge-base/how-to-deploy-nsv-kvm-imag…<https://urldefense.com/v3/__https:/www.sonicwall.com/support/knowledge-base…> Is this possible on oVirt? I probably already know the answer (no) but I want to be sure. Met vriendelijke groeten, Domien Van Rompaey Security Engineer Cloud & ICT‑infrastructure | Security Tel +32 51 76 00 00 [https://website-api.dynamate.be/public/E-mailhandtekening/DYNAMATE-handteke…]<https://urldefense.com/v3/__https:/eu.content.exclaimer.net/?url=https*3A*2…> www.dynamate.be<https://urldefense.com/v3/__https:/eu.content.exclaimer.net/?url=https*3A*2…> [https://website-api.dynamate.be/public/E-mailhandtekening/instagram-60-150.…]<https://urldefense.com/v3/__https:/eu.content.exclaimer.net/?url=https*3A*2…> [https://website-api.dynamate.be/public/E-mailhandtekening/facebook-60-150.p…]<https://urldefense.com/v3/__https:/eu.content.exclaimer.net/?url=https*3A*2…> [https://website-api.dynamate.be/public/E-mailhandtekening/linkedin-60-150-2…]<https://urldefense.com/v3/__https:/eu.content.exclaimer.net/?url=https*3A*2…>

1 0

Custom OMVF for VM
by Domien Van Rompaey 08 Jul '25

08 Jul '25

Hi We want to use the Sonicwall NSv appliance on our oVirt Cluster. Since version 7.1.1 they require the installation of a custom OMVF (UEFI Firmware). Sonicwall KB: https://www.sonicwall.com/support/knowledge-base/how-to-deploy-nsv-kvm-imag… Is this possible on oVirt? I probably already know the answer (no) but I want to be sure. Met vriendelijke groeten, Domien Van Rompaey Security Engineer Cloud & ICT‑infrastructure | Security Tel +32 51 76 00 00 [https://website-api.dynamate.be/public/E-mailhandtekening/DYNAMATE-handteke…]<https://eu.content.exclaimer.net/?url=https%3A%2F%2Fwww.christiaens.net%2F&…> www.dynamate.be<https://eu.content.exclaimer.net?url=https%3A%2F%2Fwww.dynamate.be%2F&tenan…> [https://website-api.dynamate.be/public/E-mailhandtekening/instagram-60-150.…]<https://eu.content.exclaimer.net/?url=https%3A%2F%2Fwww.instagram.com%2Fwea…> [https://website-api.dynamate.be/public/E-mailhandtekening/facebook-60-150.p…]<https://eu.content.exclaimer.net/?url=https%3A%2F%2Fwww.facebook.com%2Fwear…> [https://website-api.dynamate.be/public/E-mailhandtekening/linkedin-60-150-2…]<https://eu.content.exclaimer.net/?url=https%3A%2F%2Fwww.linkedin.com%2Fcomp…>

1 0

PsqlException at Engine
by Christoph Koehler 07 Jul '25

07 Jul '25

Dear all, Since the last host update to the latest master snapshot, we have encountered a situation where hosts with the following CPU can no longer be activated: Xeon Gold 5418Y. The error comes from a database function of the engine: "PSQLEXCEPTION: ERROR: value too long for type character varying(4000) Where: SQL statement "UPDATE vds_dynamic SET cpuflags = vcpu_flags WHERE vdsid = vvds_id" PL/pgSQL function updatecpuflags(uuid, character varying) line 3 at SQL statement" Since the update, the length of the CPU flags exceeds the datatype varchar(4000) as declared in the DB. Currently, we have no solution, and the nodes remain 'unassigned'. Does anyone know of a workaround or when this presumably known error will be fixed? Thank you and best regards, Christoph

2 3

Unable to start engine, vm status "Paused"
by mrtnz.om＠gmail.com 01 Jul '25

01 Jul '25

Hello everyone, the engine switched to paused state after attempting to place a host into maintenance, specifically when i selected the option to stop the gluster service. Glusterfs is 3 bricks [root@ovirt10 glusterfs]# gluster v heal engine info Brick 10.64.7.10:/gluster_bricks/engine/engine Status: Connected Number of entries: 0 Brick 10.64.7.11:/gluster_bricks/engine/engine Status: Connected Number of entries: 0 Brick 10.64.7.12:/gluster_bricks/engine/engine Status: Connected Number of entries: 0 --== Host ovirt11 (id: 3) status ==-- Host ID : 3 Host timestamp : 34188104 Score : 3400 Engine status : {"vm": "up", "health": "bad", "detail": "Paused", "reason": "bad vm status"} Cluster is in global maintenance The ovirt node i attempted to place in maintenance had no vms running (ovirt21) attempting to restart the engine when gluster is not running on ovirt21 results in the following error: MainThread::WARNING::2025-07-01 13:05:25,988::storage_broker::100::ovirt_hosted_engine_ha.broker.storage_broker.StorageBroker::(__init__) Can't connect vdsm storage: path to storage domain 4d31ba9e-5167-4af8-bc2a-361cce9b8278 not found in /rhev/data-center/mnt/glusterSD If i restart glusterd, and ovirt-agent and ovirt-broker on the host i put into maintenance (ovirt21), then attempt to restart the engine vm, it starts up but then is paused shortly after when glusterd service is automatically stopped on ovirt node 21. It seems the engine is stuck trying to put that host into maintenance, as soon as the engine vm starts it's paused when glusterd is stopped on ovirt21. Is there a way to safely unmount the engine gluster volume from the empty host without affecting the rest of the cluster? and/or, is there a way to stop the engine from putting that host into maintenance? Thanks! Mario M

1 0

Unable to start engine, vm status "Paused"
by mrtnz.om＠gmail.com 01 Jul '25

01 Jul '25

Hello everyone, the engine switched to paused state after attempting to place a host into maintenance, specifically when i selected the option to stop the gluster service. Glusterfs is 3 bricks [root@ovirt10 glusterfs]# gluster v heal engine info Brick 10.64.7.10:/gluster_bricks/engine/engine Status: Connected Number of entries: 0 Brick 10.64.7.11:/gluster_bricks/engine/engine Status: Connected Number of entries: 0 Brick 10.64.7.12:/gluster_bricks/engine/engine Status: Connected Number of entries: 0 --== Host ovirt11 (id: 3) status ==-- Host ID : 3 Host timestamp : 34188104 Score : 3400 Engine status : {"vm": "up", "health": "bad", "detail": "Paused", "reason": "bad vm status"} Cluster is in global maintenance The ovirt node i attempted to place in maintenance had no vms running (ovirt21) attempting to restart the engine when gluster is not running on ovirt21 results in the following error: MainThread::WARNING::2025-07-01 13:05:25,988::storage_broker::100::ovirt_hosted_engine_ha.broker.storage_broker.StorageBroker::(__init__) Can't connect vdsm storage: path to storage domain 4d31ba9e-5167-4af8-bc2a-361cce9b8278 not found in /rhev/data-center/mnt/glusterSD If i restart glusterd, and ovirt-agent and ovirt-broker on the host i put into maintenance (ovirt21), then attempt to restart the engine vm, it starts up but then is paused shortly after when glusterd service is automatically stopped on ovirt node 21. It seems the engine is stuck trying to put that host into maintenance, as soon as the engine vm starts it's paused when glusterd is stopped on ovirt21. Is there a way to safely unmount the engine gluster volume from the empty host without affecting the rest of the cluster? and/or, is there a way to stop the engine from putting that host into maintenance? Thanks! Mario M

1 0

Ovirt 4.5.5 + Veeam Unable to get path commander
by victordanco＠hotmail.com 01 Jul '25

01 Jul '25

Good morning dear community. We are experiencing a problem when taking a backup through Veeam backup. I share the logs: Engine.log 2025-06-03 16:09:52,158-05 ERROR [org.ovirt.engine.core.sso.service.SsoService] (default task-2019) [53762455] OAuthException invalid_grant: The provided authorization grant for the auth code has expired. 2025-06-03 16:09:52,160-05 ERROR [org.ovirt.engine.core.aaa.filters.SsoRestApiAuthFilter] (default task-2006) [] Cannot authenticate using authentication Headers: invalid_grant: The provided authorization grant for the auth code has expired. 2025-06-03 16:09:52,429-05 INFO [org.ovirt.engine.core.sso.service.AuthenticationService] (default task-2006) [] User admin@internal-authz with profile [internal] successfully logged in with scopes: ovirt-app-admin ovirt-app-api ovirt-app-portal ovirt-ext=auth:sequence-priority=~ ovirt-ext=revoke:revoke-all ovirt-ext=token-info:authz-search ovirt-ext=token-info:public-authz-search ovirt-ext=token-info:validate ovirt-ext=token:password-access 2025-06-03 16:09:52,518-05 INFO [org.ovirt.engine.core.bll.aaa.CreateUserSessionCommand] (default task-2006) [60f224da] Running command: CreateUserSessionCommand internal: false. 2025-06-03 16:09:52,530-05 INFO [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (default task-2006) [60f224da] EVENT_ID: USER_VDC_LOGIN(30), User admin@internal-authz connecting from '192.168.170.52' using session 'dYQlOCMThsTOInKjX8GKWUpAk/SFspOh5vuX0efK1slxzWloBoK9jGqt7TwOapfxUR1kzGg0ctoofUC7/pjeVg==' logged in. 2025-06-03 16:09:52,784-05 WARN [org.ovirt.engine.api.restapi.util.LinkHelper] (default task-2006) [] Can't find relative path for class "org.ovirt.engine.api.resource.StorageDomainVmDiskAttachmentsResource", will return null 2025-06-03 16:09:52,784-05 WARN [org.ovirt.engine.api.restapi.util.LinkHelper] (default task-2006) [] Can't find relative path for class "org.ovirt.engine.api.resource.StorageDomainVmDiskAttachmentsResource", will return null 2025-06-03 16:10:05,258-05 INFO [org.ovirt.engine.core.sso.servlets.OAuthRevokeServlet] (default task-2006) [] User admin@internal-authz with profile [internal] successfully logged out 2025-06-03 16:10:05,267-05 INFO [org.ovirt.engine.core.bll.aaa.TerminateSessionsForTokenCommand] (default task-2019) [78785b84] Running command: TerminateSessionsForTokenCommand internal: true. 2025-06-03 16:10:05,305-05 INFO [org.ovirt.engine.core.sso.servlets.OAuthRevokeServlet] (default task-2006) [] User admin@internal-authz with profile [internal] successfully logged out 2025-06-03 16:10:05,310-05 INFO [org.ovirt.engine.core.bll.aaa.TerminateSessionsForTokenCommand] (default task-2019) [4a6cfc6] Running command: TerminateSessionsForTokenCommand internal: true. 2025-06-03 16:10:28,401-05 INFO [org.ovirt.engine.core.sso.service.AuthenticationService] (default task-2006) [] User admin@internal-authz with profile [internal] successfully logged in with scopes: ovirt-app-admin ovirt-app-api ovirt-app-portal ovirt-ext=auth:sequence-priority=~ ovirt-ext=revoke:revoke-all ovirt-ext=token-info:authz-search ovirt-ext=token-info:public-authz-search ovirt-ext=token-info:validate ovirt-ext=token:password-access 2025-06-03 16:10:28,453-05 INFO [org.ovirt.engine.core.bll.aaa.CreateUserSessionCommand] (default task-2006) [152dcb43] Running command: CreateUserSessionCommand internal: false. 2025-06-03 16:10:28,466-05 INFO [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (default task-2006) [152dcb43] EVENT_ID: USER_VDC_LOGIN(30), User admin@internal-authz connecting from '192.168.170.52' using session '1WDL5jCeJNVnQCOSAJu8hd+M1IJJp5AChx3SQnpNjG+v2cZHNRYBsrQpmfhNAuB5KAHUun7HKUxbRQcs5OCalQ==' logged in. 2025-06-03 16:10:50,031-05 WARN [org.ovirt.engine.core.utils.virtiowin.VirtioWinReader] (EE-ManagedScheduledExecutorService-engineScheduledThreadPool-Thread-19) [6b0bdae7] Directory '/usr/share/virtio-win' doesn't exist. --------- Veeam Proxy: 2025-06-03 16:10:21.0156 00023 [194115] ERROR | [RhvCluster][myengine.co]: <== Request "Get" "https://myengine.co/ovirt-engine/api/vms/0a5b4e0e-477c-494a-ba43-1e6c48565a…", body: "{}" 2025-06-03 16:10:21.0156 00023 [194115] ERROR | [RhvCluster][myengine.co]: ==> Response "Get" "https://myengine.co/ovirt-engine/api/vms/0a5b4e0e-477c-494a-ba43-1e6c48565a…", "code: 401", duration: "163 msec", body: "[binary data]" Note that the user I use is admin@internal. I have read other forums that it is recommended to use admin@ovirt@internalsso or others, but veeam does not allow me to use admin@ovirt@internalsso. I appreciate your help or recommendations. Regards!

2 1

oVirt 4.5.7 - Rocky 9.4 - Update Quandry
by Clint Boggio 24 Jun '25

24 Jun '25

Good Day All - We're looking to run some updates to our oVirt environment and we're getting an error regarding selinux-policy and it's available version: Error: Problem 1: cannot install the best update candidate for package ovirt-vmconsole-1.0.9-3.el9.noarch - nothing provides selinux-policy >= 38.1.56-1.el9 needed by ovirt-vmconsole-1.0.9-3.20250519152134.git4943ea2.el9.noarch from copr:copr.fedorainfracloud.org:ovirt:ovirt-master-snapshot - nothing provides selinux-policy-base >= 38.1.56-1.el9 needed by ovirt-vmconsole-1.0.9-3.20250519152134.git4943ea2.el9.noarch from copr:copr.fedorainfracloud.org:ovirt:ovirt-master-snapshot Problem 2: package ovirt-vmconsole-host-1.0.9-3.20250519152134.git4943ea2.el9.noarch from copr:copr.fedorainfracloud.org:ovirt:ovirt-master-snapshot requires ovirt-vmconsole = 1.0.9-3.20250519152134.git4943ea2.el9, but none of the providers can be installed - cannot install the best update candidate for package ovirt-vmconsole-host-1.0.9-3.el9.noarch - nothing provides selinux-policy >= 38.1.56-1.el9 needed by ovirt-vmconsole-1.0.9-3.20250519152134.git4943ea2.el9.noarch from copr:copr.fedorainfracloud.org:ovirt:ovirt-master-snapshot - nothing provides selinux-policy-base >= 38.1.56-1.el9 needed by ovirt-vmconsole-1.0.9-3.20250519152134.git4943ea2.el9.noarch from copr:copr.fedorainfracloud.org:ovirt:ovirt-master-snapshot Problem 3: package python3.11-ovirt-imageio-client-2.5.1-0.202504080926.gitfde3ef0.el9.x86_64 from copr:copr.fedorainfracloud.org:ovirt:ovirt-master-snapshot requires python3.11-ovirt-engine-sdk4, but none of the providers can be installed Any guidance would be appreciated

1 0

Re: Network Performance Issue on oVirt Node Using Intel Platinum 8562Y+ (Sapphire Rapids) CPU
by domien.vanrompaey＠dynamate.be 24 Jun '25

24 Jun '25

Hi My issue was solved by updating the firmware on our Broadcom BCM57504 NIC (buggy firmware was 23.11.16.22). I also had to add the following option via ethtool for the NICs that were serving the bond: --offload eno12419np2 gro off --offload eno12429np3 gro off . This completely solved my problems. Maybe this can be relevant for your issue too.

1 0

Veeam - Move data over other network than ovirtmgmt
by domien.vanrompaey＠dynamate.be 18 Jun '25

18 Jun '25

Hi all I was wondering if anyone managed to get Veeam working with a dedicated backup VLAN instead of sending traffic over the ovirtmgmt interface. We have an OLVM setup (version 4.5.5-1.42.el8) with 3 hosts in 1 cluster for now. Each host has a 1Gbps NIC pair in a seperate management LAN that is only used for ovirtmgmt Each host also has a 10Gbps NIC Bond where all the VLAN's are trunked including a dedicated Backup VLAN. The hosts do not have an IP in this Backup VLAN configured (is this needed?) Each host also has 2 NICs dedicated to storage backend (iSCSI Multipathing) Our Veeam Appliance and workers are setup in the Backup VLAN and they can communicate with the backup repository that is also in this VLAN. When we take a backup, however, we see that the ovirtmgmt network is maxed out on trafic, which probably shouldn't happen. We come from a VMWare environment where the backup proxies handle all the trafic through the backup VLAN without relying on the management network and I was expecting it to work similarly Has anyone managed to get this working as I envision it? Or is the solution to simply link ovirtmgmt to the 10Gbps bond and skip the dedicated mgmt networking? I've also opened a ticket with Veeam for this, but for now I'm just getting questions from them and no real answers. If I get a useful solution from them I will also post it here

4 3

Network Performance Issue on oVirt Node Using Intel Platinum 8562Y+ (Sapphire Rapids) CPU
by dhayanands＠botvfx.com 17 Jun '25

17 Jun '25

Hi team, We’re facing a strange issue with network performance on one of our oVirt hosts and would like to get insights or guidance on possible misconfigurations or compatibility problems. Environment Details: oVirt Version: 4.5.5 Cluster CPU Type: Intel IceLake Server Family (default) Host A (Issue observed): CPU: Intel Xeon Platinum 8562Y+ (Sapphire Rapids) CPU Family/Model: Family 6 / Model 207 / Stepping 2 Host B (Working fine): CPU: Intel Xeon Gold 6354 (Ice Lake) CPU Family/Model: Family 6 / Model 106 / Stepping 6 Problem Description: We have a VM (kitsu) that was initially hosted on botovirtnode-006 (Platinum 8562Y+). The VM showed very poor download speeds (~10 Mbps), while upload was fine (~680 Mbps). After migrating the same VM to another host (botovirtnode-001 with Gold 6354 CPU), network performance drastically improved: Test On Affected Node On Healthy Node Download ~10 Mbps ~920 Mbps Upload ~680 Mbps ~760 Mbps This clearly indicates a host-specific issue, most likely related to hardware compatibility or CPU feature masking. Troubleshooting Done: Verified VM configurations (same on both hosts). Tested with speedtest-cli multiple times to confirm. Confirmed that VM virtual NIC, vNIC profile, and vSwitch settings are identical. Both hosts are part of the same cluster using the default Intel IceLake Server CPU type. Suspect the issue is due to CPU feature mismatch or unsupported stepping/model in the cluster’s CPU type.

3 3