July 2024 - Users - oVirt List Archives

Hosted engine deployment fails on Rocky Linux 9
by mpelicci＠cisco.com 18 Jul '24

18 Jul '24

Trying to do a fresh install of ovirt 4.5 on a Rocky Linux 9 host. While running the hosted-engine --deploy --4, I input all of the information. Why is it that the hosted engine is trying to reach out to the CentOS8 repos and is there a way for me to pause the deployment to modify the repo files on the vm? [ INFO ] TASK [ovirt.ovirt.engine_setup : Install oVirt Engine package] [ ERROR ] fatal: [localhost -> 192.168.1.103]: FAILED! => {"changed": false, "msg": "Failed to download metadata for repo 'centos-ceph-pacific': Cannot prepare internal mirrorlist: Curl error (6): Couldn't resolve host name for http://mirrorlist.centos.org/?release=8-stream&arch=x86_64&repo=storage-cep… [Could not resolve host: mirrorlist.centos.org]", "rc": 1, "results": []} [root@lab-node1]# cat /etc/rocky-release Rocky Linux release 9.4 (Blue Onyx)

3 2

oVirt 4.5.5 compatibility with CPU
by jakov.jelinic＠adriatic.hr 18 Jul '24

18 Jul '24

Hi, We plan to install ovirt node 4.5.5 (https://resources.ovirt.org/pub/ovirt-4.5/iso/ovirt-node-ng-installer/) on our 2 new servers, but looking at the documentation (https://www.ovirt.org/documentation/installing_ovirt_as_a_self-hosted_engin…) it seems that oVirt is not compatible with our CPU Intel Xeon Silver 4410Y 12C 150 W 2 .0 GHz processor? Can you help us?

6 6

fence_rhevm with oVirt >4.5
by justindavis＠mail.utexas.edu 18 Jul '24

18 Jul '24

Hello Folks Is the "fence_rhevm" package compatible with oVirt >4.5? We're converting our legacy RHV cluster to a new install of oVirt latest with NFS backed storage. Everything has been working smoothly with the exception of the RHV-M fencing device on this newly provisioned oVirt 4.5.7 cluster with RHEL 9.4 hosts -- I haven't been able to get authentication working between the fencing device and the manager appliance. The same configuration is working on both the legacy RHV 4.4 cluster (RHV nodes) and a test oVirt 4.5.6 cluster (RHEL 8.9 hosts). The primary difference that comes to mind between my test and new cluster is that the newest one was installed with Keycloak SSO default configs while it was disabled on the older test environment. I suspect it has something to do with dropping basic auth? Assuming this is the case, can Keycloak be removed without having to rebuild the cluster? Are there any significant drawbacks to disabling it? I've found docs for converting from AAA to Keycloak, but not the reverse. I see on the mailing list that the `ovirt-aaa-jdbc-tool` is deprecated and that Keycloak is strongly recommended moving forward -- is it possible to integrate an internal Keycloak implementation with the existing "fence_rhevm" package? The errors I'm seeing are: 401 Unauthorized This server could not verify that you are authorized to access the document requested. Either you supplied the wrong credentials (e.g., bad password), or your browser doesn't understand how to supply the credentials required. I've tried every variation of the <domain> value suggested -- including "admin@internal", "admin@ovirt@internal", "@admin@ovirt@internal-authz" Thanks in advance, Justin

2 1

New Member Introduction
by Maria Jonas 16 Jul '24

16 Jul '24

Hi everyone, I am new to this forum. I am excited to be here and to learn more about oVirt. I have just started exploring oVirt, and I have a few questions. Could someone please guide me on how to ask questions here ? Where should I post if I have a query ? Looking forward to your advice and connecting with you all. Thanks

2 1

Network issue - self-hosted engine on a single host
by jakov.jelinic＠adriatic.hr 16 Jul '24

16 Jul '24

Hi, I'm confused about installing ovirt as a self-hosted engine on a single host with local storage. I have one server which I have connected to 2 switches (which are in "Cisco stack"). I have connected the server with 2 UTP cables in master switch and 2 UTP cables in slave. I have installed on server Rocky Linux 9.4. And now i have 4 network interfaces (4 ip addresses) which i can ping from my computer. After that I installed ovirt engine 4.5.6-1.el9 with the next commands: > dnf install centos-release-ovirt45 > dnf install ovirt-engine > engine-setup Next step, I want to add a new host through my engine (Compute -> Hosts -> New). I started installing a new host. I choose my hostname to be my one and only server. After installation, I can no longer connect to my server: https://vhost01.testing.local/ovirt-engine/ because the ip address behind the dns name can no longer be pinged. Out of my 4 IP addresses, I can now ping only one from my computer. And then, of course, I change the DNS record and set it to that one-onlyworking IP address. So I think after adding the host my network is messed up. It even makes sense to me since everything is done on one host, and it seems to me that if I had separated the engine, it would not have these problems. This was my network before adding new host: ens13f0np0: connected to ens13f0np0 "Broadcom and subsidiaries BCM57454 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb" ethernet (bnxt_en), 14:23:F2:5A:B8:E4, hw, mtu 1500 ip4 default inet4 192.168.50.60/24 route4 192.168.50.0/24 metric 100 route4 default via 192.168.50.2 metric 100 inet6 fe80::1623:f2ff:fe5a:b8e4/64 route6 fe80::/64 metric 1024 ens13flnpl: connected to ens13flnpl "Broadcom and subsidiaries BCM57454 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb" ethernet (bnxt_en), 14:23:F2:5A:B8:E5, hw, mtu 1500 inet4 192.168.50.54/24 route4 192.168.50.0/24 metric 101 route4 default via 192.168.50.2 metric 101 inet6 fe80::1623:f2ff:fe5a:b8e5/64 route6 fe80::/64 metric 1024 ens13f2np2: connected to ens13f2np2 "Broadcom and subsidiaries BCM57454 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb" ethernet (bnxt_en), 14:23:F2:5A:B8:E6, hw, mtu 1500 inet4 192.168.50.78/24 route4 192.168.50.0/24 metric 102 route4 default via 192.168.50.2 metric 102 inet6 fe80::1623:f2ff:fe5a:b8e6/64 route6 fe80::/64 metric 1024 ens13f3np3: connected to ens13f3np3 "Broadcom and subsidiaries BCM57454 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb" ethernet (bnxt_en), 14:23:F2:5A:B8: E7, hw, mtu 1500 inet4 192.168.50.109/24 route4 192.168.50.0/24 metric 103 route4 default via 192.168.50.2 metric 103 inet6 fe80::1623:f2ff:fe5a:b8e7/64 route6 fe80::/64 metric 1024 lo: connected (externally) to lo "10" loopback (unknown), 00:00:00:00:00:00, sw, mtu 65536 inet4 127.0.0.1/8 inet6 ::1/128 route6:1/128 metric 256 virbro: connected (externally) to virbro "virbro" bridge, 52:54:00:11:F7:8C, sw, mtu 1500 inet4 192.168.122.1/24 route4 192.168.122.0/24 metric 0 and this is after adding new host: ens13f1np1: connected to ens13f1np1 "Broadcom and subsidiaries BCM57454 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb" ethernet (bnxt_en), 14:23:F2:5A:B8:E5, hw, mtu 1500 ip4 default inet4 192.168.50.54/24 route4 192.168.50.0/24 metric 100 route4 default via 192.168.50.2 metric 100 inet6 fe80::1623:f2ff:fe5a:b8e5/64 route6 fe80::/64 metric 1024 ovirtmgmt: connected to ovirtmgmt "ovirtmgmt" bridge, 14:23:F2:5A:B8:E4, sw, mtu 1500 inet4 192.168.50.60/24 route4 192.168.50.0/24 metric 425 inet6 fe80::1623:f2ff:fe5a:b8e4/64 route6 fe80::/64 metric 1024 ens13f2np2: connected to ens13f2np2 "Broadcom and subsidiaries BCM57454 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb" ethernet (bnxt_en), 14:23:F2:5A:B8:E6, hw, mtu 1500 inet4 192.168.50.78/24 route4 192.168.50.0/24 metric 101 route4 default via 192.168.50.2 metric 101 inet6 fe80::1623:f2ff:fe5a:b8e6/64 route6 fe80::/64 metric 1024 ens13f3np3: connected to ens13f3np3 "Broadcom and subsidiaries BCM57454 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb" ethernet (bnxt_en), 14:23:F2:5A:B8:E7, hw, mtu 1500 inet4 192.168.50.109/24 route4 192.168.50.0/24 metric 102 route4 default via 192.168.50.2 metric 102 inet6 fe80::1623:f2ff:fe5a:b8e7/64 route6 fe80::/64 metric 1024 ens13f0np0: connected to ens13f0np0 "Broadcom and subsidiaries BCM57454 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb" ethernet (bnxt_en), 14:23:F2:5A:B8:E4, hw, mtu 1500 master ovirtmgmt lo: connected (externally) to lo "lo" loopback (unknown), 00:00:00:00:00:00, sw, mtu 65536 inet4 127.0.0.1/8 inet6 ::1/128 route6 ::1/128 metric 256 br-int: disconnected "br-int" ovs-interface (openvswitch), 12:47:1C:18:8F:E2, sw, mtu 1500 Can you plese help?

2 1

HostedEngine migration fail
by doboscsongor＠gmail.com 16 Jul '24

16 Jul '24

Hi, I have two hosts in one cluster, where the hosts are exactly the same physical servers. The migration of virtual machines works fine, but I can't migrate the hosted engine VM. When I try to migrate it to the second host/server, the "Destination host" field in the migration pop-up window shows "No available host to migrate VMs to." However, if I try to migrate any other running VMs, the field becomes active and I can choose the second host/server. Screenshot: https://i.postimg.cc/YCnddhs1/ovirt.jpg Server software information: - OS Version: RHEL - 9.1.2206.0 - 23.el9 - OS Description: oVirt Node 4.5.5 - Kernel Version: 5.14.0 - 388.el9.x86_64 - KVM Version: 8.1.0 - 4.el9 - LIBVIRT Version: libvirt-9.5.0-6.el9 - VDSM Version: vdsm-4.50.5.1-1.el9 Thanks in advance!

2 1

[Self-hosted engine] Change port for web portal
by junhoyu＠beamworks.co.kr 16 Jul '24

16 Jul '24

Hi everyone, I've installed oVirt (specifically Oracle Linux Virtualization Manager v.4.5) on my server using the Self-Hosted Deployment method. However, my server is in an environment where using well-known ports is restricted. Therefore, I need to reconfigure the ports for the web portal, which currently uses ports 80 and 443, so that I can access the web portal through the browser. I tried the following two methods: === Method 1 === During installation, I prepared an answer file (/root/answers.conf) with the following configurations: OVESETUP_CONFIG/publicHttpPort=int:2416 OVESETUP_CONFIG/publicHttpsPort=int:5281 OVESETUP_CONFIG/httpPort=int:2416 OVESETUP_CONFIG/httpsPort=int:5281 Then, I executed 'hosted-engine --deploy --4 --config-append=/root/answers.conf'. However, this did not change the ports. === Method 2 === After installation, I modified the firewalld service files (ovirt-http.xml, ovirt-https.xml), httpd configuration files (httpd.conf, conf.d/ssl.conf), and engine configuration files (engine.conf.d/10-setup-portocols.conf, 11-setup-sso.conf). This worked, but the 'hosted-engine --vm-status' command contains the following issue: Engine status : {"vm": "up", "health": "bad", "detail": "Up", "reason": "failed liveliness check"} It seems to be caused by the health check to the web portal. (https://{engine_fqdn}/ovirt-engine/services/health -> the health-check process does not modify the port in this URL) ====== So, I will solve it with aspects related to the two questions below. 1) Is there a way to change the engine health-check URL during or after installation? 2) Is there a way to configure ports globally related to oVirt? I hope you guys can provide solutions. Thanks. Best regards, Junho Yu

1 0

lvm.log is growing
by dev＠eazis.com 16 Jul '24

16 Jul '24

Hi, The file /tmp/lvm.log is growing every day. I need to clear the file once in the two weeks. Can anyone tell me how i can rotate this log file? Kind regards, Gerjan

1 0

Start VM with option -machine smm=off
by Kalil de A. Carvalho 16 Jul '24

16 Jul '24

Hello all. I'm trying to workaround the problem reported here: https://gitlab.com/qemu-project/qemu/-/issues/1198 but I did not find out how to start up the VM with the option -machine smm=off like sead on the lsit above. I guess that it is possible to run the command using the qemu command but I think that I will lose the management of it. It is possible to startup up a VM with this option using the oVirt GUI ou virsh? Best regards -- Atenciosamente, Kalil de A. Carvalho

3 2

New oVirt release ???
by Diggy 15 Jul '24

15 Jul '24

Any movement towards a new oVirt release now that CentOS Stream 8 is officially at EOL and the current engine is based on it? Do I see oVirt v4.6 in my near future? I'm looking for a reason to build a fresh oVirt environment and migrate our 100+ VMs to it. :) ---

6 10