- Users - oVirt List Archives

rhv-log-collector-analyzer available missing?
by Juhani Rautiainen 20 Apr '21

20 Apr '21

Hi! I'm trying to upgrade SHE from 4.3->4.4. Instructions from https://www.ovirt.org/documentation/upgrade_guide/#SHE_Upgrading_from_4-3 has step "4.2 Analyzing the Environment". My installation doesn't bring the tool rhv-log-collector-analyzer although wiki has a notice that it has been available since v4.2.5? [root@ovirtmgr ~]# yum install rhv-log-collector-analyzer Loaded plugins: fastestmirror, versionlock Determining fastest mirrors ovirt-4.3-epel/x86_64/metalink | 34 kB 00:00:00 * base: mirror.hosthink.net * extras: mirror.hosthink.net * ovirt-4.3: resources.ovirt.org * ovirt-4.3-epel: mirrors.nxthost.com * updates: mirror.hosthink.net base | 3.6 kB 00:00:00 centos-sclo-rh-release | 3.0 kB 00:00:00 extras | 2.9 kB 00:00:00 ovirt-4.3 | 3.0 kB 00:00:00 ovirt-4.3-centos-gluster6 | 3.0 kB 00:00:00 ovirt-4.3-centos-opstools | 2.9 kB 00:00:00 ovirt-4.3-centos-ovirt-common | 3.0 kB 00:00:00 ovirt-4.3-centos-ovirt43 | 2.9 kB 00:00:00 ovirt-4.3-centos-qemu-ev | 3.0 kB 00:00:00 ovirt-4.3-epel | 4.7 kB 00:00:00 ovirt-4.3-virtio-win-latest | 3.0 kB 00:00:00 sac-gluster-ansible | 3.3 kB 00:00:00 updates | 2.9 kB 00:00:00 (1/9): extras/7/x86_64/primary_db | 232 kB 00:00:00 (2/9): ovirt-4.3-centos-gluster6/x86_64/primary_db | 120 kB 00:00:00 (3/9): ovirt-4.3-epel/x86_64/group_gz | 96 kB 00:00:00 (4/9): base/7/x86_64/primary_db | 6.1 MB 00:00:00 (5/9): ovirt-4.3-epel/x86_64/updateinfo | 1.0 MB 00:00:00 (6/9): ovirt-4.3-epel/x86_64/primary_db | 6.9 MB 00:00:00 (7/9): centos-sclo-rh-release/x86_64/primary_db | 2.9 MB 00:00:00 (8/9): sac-gluster-ansible/x86_64/primary_db | 12 kB 00:00:00 (9/9): updates/7/x86_64/primary_db | 7.1 MB 00:00:00 No package rhv-log-collector-analyzer available. Error: Nothing to do Are we missing a repo or is this just copy/paste error from RHV docs and this step shouldn't even be in the oVirt docs? Thanks, Juhani -- Juhani Rautiainen jrauti(a)iki.fi

3 3

Re: FreeBSD 13 and virtio
by Thomas Hoberg 20 Apr '21

20 Apr '21

I've just given it a try: works just fine with me. But I did notice, that I chose virtio-scsi when I created the disk, don't now if that makes any difference, but as an old-timer, I still got "SCSI" ingrained as "better than ATA". Chose FreeBSD 9.2 x64 as OS type while creating the VM (nothing closer appears in the selection box), again don't know if there are better choices. iperf3 shows slightly lower performance than a typical Linux guest, still rather close to 10Gbit/s on my network from the FreeBSD13 guest to a host "next-door", around 40Gbit/s VM to host. Verfied usage of vtnet0/VirtIO NIC and vtscsi0/VirtIO SCSI via dmesg.

2 1

Is it possible to upgrade 3 node HCI from 4.3 to 4.4?
by Jayme 20 Apr '21

20 Apr '21

I have a fairly stock three node HCI setup running oVirt 4.3.9. The hosts are oVirt node. I'm using GlusterFS storage for the self hosted engine and for some VMs. I also have some other VMs running from an external NFS storage domain. Is it possible for me to upgrade this environment to 4.4 while keeping some of the VMs on GlusterFS storage domains running? Does the upgrade require an additional physical host? I'm reading through the upgrade guide and it's not very clear to me how the engine is reinstalled on CentOS8. Do you need a 4th host, or do you take one of the three existing hosts down, wipe it and install ovirt node 4.4? Would it be easier for me to just move all VMs to my NFS storage domain, wipe all hosts and deploy a brand new 4.4 HCI cluster then import the NFS storage domain? Will the 4.3 VMs on the NFS storage domain import/run properly into the new 4.4 deployment or are there any other considerations?

3 4

intel_iommu=on kvm-intel.nested=1 deactivates rp_filter kernel option
by Nathanaël Blanchet 20 Apr '21

20 Apr '21

Hello, I have two kind of hosts: * some with default ovirt node 4.4 kernel settings * some with custom kernel settings including intel_iommu=on kvm-intel.nested=1 I can't open vm console from the second category host when binding from a different vlan because the host is unreacheable But if I set sysctl -w net.ipv4.conf.all.rp_filter=2, I can bind the host and open a vm console. I didn't test if this behaviour is because of Hostdev Passthrough & SR-IOV or Nested Virtualization. Is it an expected behaviour or a bug? -- Nathanaël Blanchet Supervision réseau SIRE 227 avenue Professeur-Jean-Louis-Viala 34193 MONTPELLIER CEDEX 5 Tél. 33 (0)4 67 54 84 55 Fax 33 (0)4 67 54 84 14 blanchet(a)abes.fr

2 1

Two selfhosted engines
by wilrikvdvliert＠hotmail.com 20 Apr '21

20 Apr '21

Dear forum, If I understand correctly, I need shared storage for the self-hosted engine to make it highly available. But is it possible to create 2 self-hosted engines on 2 hosts and make it high available this way? Kind regard

3 2

ova import
by eevans＠digitaldatatechs.com 20 Apr '21

20 Apr '21

I am running Ovirt 4.3 and exported machines as ova which includes disks. When I re import them. the disks are thin provisioned and the vm will not boot. Any idea how to get the import to preallocate disks? I have vms that will not boot. It states no bootable disk is available. It seems the virt2vm isn't working correctly. ANy help would be appreciated. Eric

4 5

[Help] The engine is not stable on HostedEngine with GlusterFS Herperverged deployment
by mengz.you＠outlook.com 20 Apr '21

20 Apr '21

Hi, I deployed a oVirt (4.3.10) cluster with HostedEngine and GlusterFS volumes (engine, vmstore, data), the glusterfs cluster on node1/node2/node3, and the engine vm can be running on those 3 nodes. Then I added a 4th nodes into cluster. But, when I operates on Eninge Web Portal, it's always reports 503 error, then I checked `hsoted-engine --vm-status`, see below: ``` [root@vhost1 ~]# hosted-engine –vm-status –== Host vhost1.yhmk.lan (id: 1) status ==– conf_on_shared_storage : True Status up-to-date : True Hostname : vhost1.<span style=”background-color: rgb(255, 255, 255); color: rgb(51, 51, 51);”>alatest</span>.lan Host ID : 1 Engine status : {“reason”: “bad vm status”, “health”: “bad”, “vm”: “down_unexpected”, “detail”: “Down”} Score : 0 stopped : False Local maintenance : False crc32 : 1f25baff local_conf_timestamp : 1253650 Host timestamp : 1253649 Extra metadata (valid at timestamp): metadata_parse_version=1 metadata_feature_version=1 timestamp=1253649 (Thu Apr 8 08:05:48 2021) host-id=1 score=0 vm_conf_refresh_time=1253650 (Thu Apr 8 08:05:48 2021) conf_on_shared_storage=True maintenance=False state=EngineUnexpectedlyDown stopped=False timeout=Thu Jan 15 20:23:29 1970 –== Host vhost2.yhmk.lan (id: 2) status ==– conf_on_shared_storage : True Status up-to-date : True Hostname : vhost2.<span style=”background-color: rgb(255, 255, 255); color: rgb(51, 51, 51);”>alatest</span>.lan Host ID : 2 Engine status : {“reason”: “vm not running on this host”, “health”: “bad”, “vm”: “down_unexpected”, “detail”: “unknown”} Score : 3400 stopped : False Local maintenance : False crc32 : 539fc30c local_conf_timestamp : 1253343 Host timestamp : 1253343 Extra metadata (valid at timestamp): metadata_parse_version=1 metadata_feature_version=1 timestamp=1253343 (Thu Apr 8 08:05:46 2021) host-id=2 score=3400 vm_conf_refresh_time=1253343 (Thu Apr 8 08:05:46 2021) conf_on_shared_storage=True maintenance=False state=EngineDown stopped=False –== Host vhost3.yhmk.lan (id: 3) status ==– conf_on_shared_storage : True Status up-to-date : True Hostname : vhost3.alatest.lan Host ID : 3 Engine status : {“reason”: “bad vm status”, “health”: “bad”, “vm”: “up”, “detail”: “Powering up”} Score : 3400 stopped : False Local maintenance : False crc32 : 4072e0b8 local_conf_timestamp : 1252345 Host timestamp : 1252345 Extra metadata (valid at timestamp): metadata_parse_version=1 metadata_feature_version=1 timestamp=1252345 (Thu Apr 8 08:05:42 2021) host-id=3 score=3400 vm_conf_refresh_time=1252345 (Thu Apr 8 08:05:42 2021) conf_on_shared_storage=True maintenance=False state=EngineStarting stopped=False ``` Then, wait a moment, can access web portal again, and check the hosts status, alway reports one or more hosts with label `unavaiable as HA score`, but it will dispear later. And I found, sometimes the engine vm will migrated to another nodes whill this problem occur. So, seems the HostedEngine is not stable, always occur this problem, could you please help me with this ? Thanks!

3 8

Re: ovirt-hosted-engine-cleanup
by Thomas Hoberg 20 Apr '21

20 Apr '21

ovirt-hosted-engine-cleanup will only operate on the host you run it on. In a cluster that might have side-effects, but as a rule it will try to undo all configuration settings that had a Linux host become an HCI member or just a host under oVirt management. While the GUI will try to do the same with Ansible only, the script seems to do a little more and be more thorough. My guess is that it's somewhat of a leftover from the older 'all scripts' legacy of oVirt, before they re-engineered much of it to fit Ansible. It's not perfect, however. I have come across things that it would not undo, and that were extremely difficult to find (a really clean re-install would then help). There was an issue around keys and certificates for KVM that I only remember as a nightmare and more recently I had an issue with it's ansible scripts inserting an LVM device filter specific for a UID that the HCI setup wizard had created, which then precluded it ever working again after running the *cleanup scripts. All I got was "device excluded by a filter" and only much later I found the added line in /etc/lvm.conf not undone, which caused this nail-biter. But generally that message "Is hosted engine setup finished?" mostly indicates that the four major daemons for HCI still have issues. It all starts with glusterd: If you just added a host in HCI with storage, that would like to be properly healed. AFAIK that doesn't actually preclude already using the machine to host VMs, but in the interest of keeping your head clear, I'd suggest concentrating on getting the storage all clean, with gluster happily showing connectivity all around (gluster volume status all etc.) and 'gluster volume heal <volume> [info]' giving good results. Then you need to check on the services ovirt-ha-broker and ovirt-ha-agent as well as vdsmd via 'servicectl status <servicename>' to see what they are complaining about. I typically restart them in this order to iron out the bumps. As long as you give them a bit of time, restarting them periodically doesn't seem to do any harm, but helps them recover from hangs. And then you simply need patience: Things don't happen immediately in oVirt, because few commands ever intervene into anything directly. Instead they set flags here and there which then a myriad of interlocking state machine gears will pick up and respond to, to do things in the proper order and by avoiding the race conditions manual intervention can so easily cause. Even the fastest systems won't give results in seconds. So have a coffee, better yet, some herbal tea, and then look again. There is a good chance oVirt will eventually sort it out (but restarting those daemons really can help things along, as will looking at those error messages and their potential cause).

1 0

Introduction & general question about oVirt
by Nicolas Kovacs 20 Apr '21

20 Apr '21

Hi, I'm a 53-year old Austrian living in Montpezat, a small village in South France. I'm an IT professional with a focus on Linux and free software, and I've been a Linux user since Slackware 7.1. I'm doing web & mail hosting for myself and several small structures like our local school and a handful of local companies. Up until recently these hostings have happened on "bare metal" root servers using CentOS 7. One main server is hosting most of the stuff: WordPress sites, one OwnCloud instance, Dolibarr management software, GEPI learning platform, Postfix/Dovecot mail server, Roundcube webmail, etc. This setup has become increasingly problematic to manage, since applications have more and more specific requirements, like different versions of PHP and corresponding modules. So I decided to split everything up nicely into a series of virtual machines, each one with a nicely tailored setup. I have a couple of sandbox servers, one public and one local, running Oracle Linux 7 (a RHEL clone like CentOS). I played around with it, and KVM-based virtualization already works quite nicely. While looking for documentation, I stumbled over oVirt, which I didn't even know existed until last week. Before I dive head first into it, I'd be curious to know a few general things. 1. Would it be overkill for a small structure like mine? 2. Will I be able to do HA on a series of modest KVM-capable root servers even if they are located in different datacenters across different countries? 3. One problem I couldn't resolve using a bone-headed keep-it-simple KVM setup is backup. For my bare-metal servers I've been using incremental backups using Rsnapshot for years. Here's a blog article I wrote on the subject: https://blog.microlinux.fr/rsnapshot-centos-7/ Unfortunately I can't use this approach with huge QCOW images, at least not without jumping through burning loops. Is there an easy way to perform remote incremental backups with oVirt? BTW, I took a peek at Proxmox and Ceph, but I admit I'm a die-hard RHEL-clone user. Cheers from the sunny South of France, Niki -- Microlinux - Solutions informatiques durables 7, place de l'église - 30730 Montpezat Site : https://www.microlinux.fr Blog : https://blog.microlinux.fr Mail : info(a)microlinux.fr Tél. : 04 66 63 10 32 Mob. : 06 51 80 12 12

4 3

Upgrading single host ovirt "clusters" to v4.5 via command line?
by Gilboa Davara 19 Apr '21

19 Apr '21

Hello all, As the title suggests I have a couple of single-host hyper converged setups, mostly used for testing. As I cannot use the GUI cluster upgrade method on a single host (it requires me to reboot the host, which will require me to shutdown the hosted engine), is there any method to upgrade the cluster / data-center via command line in maintenance=global mode? Beyond that, can I use the same method on my production setups? (multi-node, gluster based setups). Thanks, Gilboa

1 0