'python-nftables' failed. There is a big problem with Firewalld and openvswitch on Centos 8.3
by Сергей Шопин
I have deployed a new cluster on versions:
3 Hosts: Centos 8.3 and last update
Ovirt: 4.4.4.7-1.el8
Cluster properties: switch - OVS; Firewall - Firewalld
During the initial deployment of the host in such a cluster, everything is successfully installed, all virtual networks and bridges are created, and the host is activated.
But after that, Firewalld does not work correctly, for example, when executing the command
firewall-cmd --reload
Issued by:
Error: COMMAND_FAILED: 'python-nftables' failed:
JSON blob:
{"nftables": [{"metainfo": {"json_schema_version": 1}}, {"add": {"table": {"family": "inet", "name": "firewalld_policy_drop"}}}, {"add": {"chain": {"family": "inet", "table": "firewalld_policy_drop", "name": "filter_input", "type": "filter", "hook": "input", "prio": 9, "policy": "drop"}}}, {"add": {"chain": {"family": "inet", "table": "firewalld_policy_drop", "name": "filter_forward", "type": "filter", "hook": "forward", "prio": 9, "policy": "drop"}}}, {"add": {"chain": {"family": "inet", "table": "firewalld_policy_drop", "name": "filter_output", "type": "filter", "hook": "output", "prio": 9, "policy": "drop"}}}, {"add": {"rule": {"family": "inet", "table": "firewalld_policy_drop", "chain": "filter_input", "expr": [{"match": {"left": {"ct": {"key": "state"}}, "op": "in", "right": {"set": ["established", "related"]}}}, {"accept": null}]}}}, {"add": {"rule": {"family": "inet", "table": "firewalld_policy_drop", "chain": "filter_forward", "expr": [{"match": {"left": {"ct": {"key": "state"}}, "op": "in", "right": {"set": ["established", "related"]}}}, {"accept": null}]}}}, {"add": {"rule": {"family": "inet", "table": "firewalld_policy_drop", "chain": "filter_output", "expr": [{"match": {"left": {"ct": {"key": "state"}}, "op": "in", "right": {"set": ["established", "related"]}}}, {"accept": null}]}}}]}
Now all ansible-playbook ovirt starts to execute uncorrectly
For example this error is thrown every time I try to reinstall the host:
VDSM SMnode02 command CollectVdsNetworkDataAfterInstallationVDS failed: Internal JSON-RPC error: {'reason': ’management’}
I understand that ovirt correctly executes scripts, and this problem is related to the applications Firewalld, openvswitch, NetworkManager.
Is there some workaround so that at least ovirt can re-install hosts?
3 years, 2 months
Unable to ping host from VM's or VM's from host
by David Johnson
I have a brand new Ovirt 4.5 cluster on centos 8 and brand new hardware.
I have a set of VM's, some of whom I have been able to connect to in the
past, others that I have never been able to connect to other than with the
SPICE console ui.
Right now, I am unable to connect from outside my cluster to any of the
VPN's other than through the SPICE console.
Host IP: 192.168.2.188 Physical Address
Example VM IP: 10.21.10.101 with vLAN tag
Works:
* I can connect between vm's on the same subnet on the same VM.
When logged into 10.21.10.101 I can ping 10.21.10.100
Doesn't work:
* I cannot connect from VM's to the host.
* I cannot connect from the host to VM's
PING 10.21.10.100 (10.21.10.100) 56(84) bytes of data.
From <<internet gateway>> icmp_seq=1 Destination Host Unreachable
* I cannot connect from the VM's to the internet gateway
* I cannot connect to the VM's from other systems on the physical network
"Destination Host Unreachable"
To me it looks like the virtual bridge is misconfigured, but I can't figure
out how to reconfigure it.
Any assistance would be greatly appreciated.
Regards,
David Johnson
3 years, 2 months
oVirt initramfs regeneration
by Robert Tongue
Is it possible to force an initramfs regeneration in oVirt Node 4.4.4? I am doing some advanced partitioning and cannot seem to figure out how to properly do that if it's even possible. Thanks.
-phunyguy
3 years, 2 months
Geographically-distrbuted Ovirt Cluster?
by Derek Atkins
Hi,
Right now I run a 1-server hyperconverged hosted-engine deployment. Other
than requiring systemwide downtime to perform host maintenance, it has
worked fairly well for me over the past few years. However, after a
couple power and upstream-network outages, there have been questions about
the ability to "distribute the load", so to speak. Considering what our
workload is (main git repo, wiki, email list + archives, etc), it can't
easily be distributed by a periodic resync and DNS round robin like a
static web site.
To solve this distribution problem I was wondering if there might be some
way to deploy a handful of physical servers in different geographic
locations that work together to create an HE environment for guest VMs?
My initial fear would be that it would require dedicated 1Gb (or higher)
between the sites to move data? Let's say we do NOT have that level of
connectivity.
I saw something about geographic replication in Gluster? But that would
seem to be more about replicating a local cluster to a remote cluster?
I suspect the answer is: no, a cluster must all be physically co-located.
But I figured I would ask the experts.
Thanks all for your insights.
-derek
--
Derek Atkins 617-623-3745
derek(a)ihtfp.com www.ihtfp.com
Computer and Internet Security Consultant
3 years, 2 months
CentOS Stream support
by Michal Skrivanek
Hi all,
we would like to ask about interest in community about oVirt moving to CentOS Stream.
There were some requests before but it’s hard to see how many people would really like to see that.
With CentOS releases lagging behind RHEL for months it’s interesting to consider moving to CentOS Stream as it is much more up to date and allows us to fix bugs faster, with less workarounds and overhead for maintaining old code. E.g. our current integration tests do not really pass on CentOS 8.1 and we can’t really do much about that other than wait for more up to date packages. It would also bring us closer to make oVirt run smoothly on RHEL as that is also much closer to Stream than it is to outdated CentOS.
So..would you like us to support CentOS Stream?
We don’t really have capacity to run 3 different platforms, would you still want oVirt to support CentOS Stream if it means “less support” for regular CentOS?
There are some concerns about Stream being a bit less stable, do you share those concerns?
Thank you for your comments,
michal
3 years, 2 months
ovirt node based on centos 8 stream
by Nathanaël Blanchet
Hi all,
I project to upgrade from 4.3 to 4.4 hosts in the next few days, and I
wonder if ovirt node based on centos 8.3 will be upgradable to ovirt
node based on centos stream.
If not, I will wait to upgrade directly to ovirt node centos stream
based when available.
--
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
3 years, 2 months
Re: VM templates
by Strahil Nikolov
First of all ,
verify the gluster volume options (gluster volume info <VOLNAME> ;
gluster volume status <VOLNAME>).When you use HCI, ovirt sets up a lot
of optimized options in order to gain the maximum of the Gluster
storage.
Best Regards,Strahil Nikolov
В 15:03 +0000 на 25.01.2021 (пн), Robert Tongue написа:
> Hello,
>
>
>
> Another weird issue over here. I have the latest oVirt running
> inside VMware Vcenter, as a proof of concept/testing platform.
> Things are working well finally, for the most part, however I am
> noticing strange behavior with templates, and deployed VMs
> from that template. Let me explain:
>
>
>
> I created a basic Ubuntu Server VM, captured that VM as a template,
> then deployed 4 VMs from that template. The deployment went fine;
> however I can only start 3 of the 4 VMs. If I shut one down one of
> the 3 that I started, I can then start the other
> one that refused to start, then the one I JUST shut down will then
> refuse to start. The error is:
>
>
>
>
> VM test3 is down with error. Exit message: Bad volume specification
> {'device': 'disk', 'type': 'disk', 'diskType': 'file', 'specParams':
> {}, 'alias': 'ua-2dc7fbff-da30-485d-891f-03a0ed60fd0a', 'address':
> {'bus': '0', 'controller': '0', 'unit': '0', 'type':
> 'drive', 'target': '0'}, 'domainID': '804c6a0c-b246-4ccc-b3ab-
> dd4ceb819cea', 'imageID': '2dc7fbff-da30-485d-891f-03a0ed60fd0a',
> 'poolID': '3208bbce-5e04-11eb-9313-00163e281c6d', 'volumeID':
> 'f514ab22-07ae-40e4-9146-1041d78553fd', 'path': '/rhev/data-
> center/3208bbce-5e04-11eb-9313-00163e281c6d/804c6a0c-b246-4ccc-b3ab-
> dd4ceb819cea/images/2dc7fbff-da30-485d-891f-03a0ed60fd0a/f514ab22-
> 07ae-40e4-9146-1041d78553fd',
> 'discard': True, 'format': 'cow', 'propagateErrors': 'off', 'cache':
> 'none', 'iface': 'scsi', 'name': 'sda', 'bootOrder': '1', 'serial':
> '2dc7fbff-da30-485d-891f-03a0ed60fd0a', 'index': 0, 'reqsize': '0',
> 'truesize': '2882392576', 'apparentsize': '3435134976'}.
>
>
>
> The underlying storage is GlusterFS, self-managed outside of oVirt.
>
>
>
>
>
> I can provide any logs needed, please let me know which. Thanks in
> advance.
>
>
>
>
>
> _______________________________________________Users mailing list --
> users(a)ovirt.org
> To unsubscribe send an email to users-leave(a)ovirt.org
> Privacy Statement: https://www.ovirt.org/privacy-policy.html
> oVirt Code of Conduct:
> https://www.ovirt.org/community/about/community-guidelines/
> List Archives:
> https://lists.ovirt.org/archives/list/users@ovirt.org/message/TZMAQ56NIQK...
3 years, 2 months
HCI Deployment via Ansible
by penguin pages
I am trying to teach myself new and better means to deploy and control oVirt and HCI deployment.
Does anyone have notes, experience and or examples.. website etc.. where someone managed deployment and admin of HCI deployment (CentOS + Cockpit + gluster + oVirt) and instead of using shell and manaual / gui deployment, did this "Infrastructure as code"
Ex Step 1:
1) Foundation with kickstart. three nodes with "minimal install" all with ssh keyless password login for user "ansible"
2) Keys exchanged between nodes for keyless login for root and ansible user
3) Using ansible publish HCI deployment playbook.
-> Define drive for each host to use
-> Define name and host / ip of engine
-> install HCI engine with gluster
-> Add SSH keys for ansible and access so future oVirt interaction can be done via ansible
I am just starting to research this.. but hoping others have blazed this path and I can leverage their process.
Concerns:
1) How to get HCI to deploy only engine layer on (in my case... 512GB SSD .. with VDO and avoid destroying VMS not related to engine).. aka if I have to repair engine again will it keep nuking the volumes "data" and "vmstore" volumes which I think the HCI deployment wants to do if it has to redeploy fresh.
2) If above is always to be an issue .. can I make default VM deployment be a dedicated gluster volume (my case is 1TB SSD with VDO in each server) such that rebuild of engine layer can just slurp back in the VMs from that volume.
3 years, 2 months
Performance issue to retrieve VM's after upgrading engine from 4.2 to 4.3
by Jonas Lindholm
This is a weird issue.
Before upgrading it took 8-9 seconds to retrieve all VM's, 1230, using /ovirt-engine/api/vms URI. After the upgrade it now takes over 2 minutes.
Query other items such as hosts, clusters, data centers etc is fast and the same speed as prior the upgrade from 4.2 to 4.3
I don't know if this is DB related or jboss related but any clue where to start to dig deeper is welcome.
I also built a new VM and restored the DB's and ran engine-setup, but that new engine also gets more or less exactly the same response times for the VM's using the restAPI.
I have another engine with same version 4.3.10.4-1, That only have 261 VMs but those are retrieved within 1-2 seconds only so I don't understand how that engine is so much faster than the one I upgraded.
The UI also has the same long response times to show VM's from Compute -> Virtual Machines
Any idea what could cause this?
3 years, 2 months
