February 2024 - openstack-discuss

[publiccloud-sig] Reminder - next meeting February 21st - 0800 UTC
by Felix Kronlage-Dammers 20 Feb '24

20 Feb '24

Hi everyone, This Wednesday (February 21st) the next meeting of the public cloud sig is going to happen. Since the majority of participants in the sessions started to like the videocalls, we’ll be doing this as a videocall again. We’ll be meeting at 0800 UTC here: https://conf.scs.koeln:8443/OIF-public-cloud-sig See also here for all other details: https://wiki.openstack.org/wiki/PublicCloudSIG We’ve had a lively discussion couple of weeks ago and Artem (gtema) did make a good head start in regards to sketching out the discovery service. I’m fairly certain that this will be part of tomorrow’s meeting ;) See you tomorrow! felix -- Felix Kronlage-Dammers Product Owner IaaS & Operations Sovereign Cloud Stack Sovereign Cloud Stack — standardized, built and operated by many Ein Projekt der Open Source Business Alliance - Bundesverband für digitale Souveränität e.V. Tel.: +49-30-206539-205 | Matrix: @fkronlage:matrix.org | fkr(a)osb-alliance.com

1 0

kolla ansible all in one VM instance creation fails
by KK CHN 19 Feb '24

19 Feb '24

List 1. I followed the tutorial https://kifarunix.com/deploy-all-in-one-openstack-with- kolla-ansible-on-ubuntu-22-04/?expand_article=1 and able to get things up and running upto horizon dashboard login.( with the runinit once script ) which brings the default cirros image and default flavors and default networks. 2. When I tried to create a cirros instance either through the horizon dashboard or through openstack CLI the instance not up and running resulting in error . 3. NOTE: This kolla-ansible-all-in-one I am trying on a VM( Ubuntu22.04 VM deployed on ESXi 7.0 host). 4. Do I need to edit or enable something on the ESXi host machine or the Ubuntu base VM where I am trying the all-in-one installation ? ERROR as follows 2024-02-19 22:14:17.881 7 ERROR nova.compute.manager [instance: ca98a487-8214-4b41-bdbc-6eed0a9390cf] libvirt.libvirtError: unsupported configuration: Emulator '/usr/bin/qemu-system-x86_64' does not support virt type 'kvm' 2024-02-19 22:14:17.881 7 ERROR nova.compute.manager [instance: ca98a487-8214-4b41-bdbc-6eed0a9390cf] 2024-02-19 22:14:17.885 7 INFO os_vif [None req-133b5da1-40ad-45d4-af54-7e4fa299aeed 76b826e348324acb8ebf005721c0213a e83c8d350df24c51851c44bf6d6fc7d5 - - default default] Successfully unplugged vif VIFBridge(active=False,address=fa:16:3e:cc:95:02,bridge_name='qbr40ec669f-fc',has_traffic_filtering=True,id=40ec669f-fc19-4c1e-af03-5180819758a6,network=Network(db91c3c3-d0ca-41d1-9497-6511e79f2b10),plugin='ovs',port_profile=VIFPortProfileOpenVSwitch,preserve_on_delete=False,vif_name='tap40ec669f-fc') 2024-02-19 22:14:18.176 7 INFO nova.compute.manager [None req-133b5da1-40ad-45d4-af54-7e4fa299aeed 76b826e348324acb8ebf005721c0213a e83c8d350df24c51851c44bf6d6fc7d5 - - default default] [instance: ca98a487-8214-4b41-bdbc-6eed0a9390cf] Took 0.29 seconds to deallocate network for instance. 2024-02-19 22:14:18.280 7 INFO nova.scheduler.client.report [None req-133b5da1-40ad-45d4-af54-7e4fa299aeed 76b826e348324acb8ebf005721c0213a e83c8d350df24c51851c44bf6d6fc7d5 - - default default] Deleted allocations for instance ca98a487-8214-4b41-bdbc-6eed0a9390cf 2024-02-19 22:14:20.426 7 WARNING nova.virt.libvirt.driver [None req-00027f3f-4666-4f97-a672-085105be0946 - - - - - -] This host appears to have multiple sockets per NUMA node. The `socket` PCI NUMA affinity will not be supported. 2024-02-19 22:15:20.419 7 WARNING nova.virt.libvirt.driver [None req-00027f3f-4666-4f97-a672-085105be0946 - - - - - -] This host appears to have multiple sockets per NUMA node. The `socket` PCI NUMA affinity will not be supported. root@ubuntu-VM:/home/ubuntu# Greetings Krish

1 1

[tc] Next Meeting Tomorrow, Feb 20, 1800 UTC
by Jay Faulkner 19 Feb '24

19 Feb '24

Hi all, the next weekly meeting of the OpenStack Technical Committee is tomorrow, Feb 20, at 1800 UTC. The proposed agenda is: Agenda - Roll call - Follow up on tracked action items - Gate health check - Implementation of Unmaintained branch statuses - Testing runtime for 2024.2 release (gmann) - https://review.opendev.org/c/openstack/governance/+/908862 - 2024.1 TC Tracker - https://etherpad.opendev.org/p/tc-2024.1-tracker - Open Discussion and Reviews Thanks, Jay Faulkner OpenStack TC Chair

1 0

Request for support for the openstack trial
by Richard Buregeya 19 Feb '24

19 Feb '24

Hello team, My nova-compute is in disable mode and here are the logs and nova-compute is down... 2024-02-19 10:49:43.131 7 ERROR oslo_service.service File "/usr/lib/python2.7/site-packages/nova/virt/libvirt/driver.py", line 638, in init_host 2024-02-19 10:49:43.131 7 ERROR oslo_service.service self._do_quality_warnings() 2024-02-19 10:49:43.131 7 ERROR oslo_service.service File "/usr/lib/python2.7/site-packages/nova/virt/libvirt/driver.py", line 604, in _do_quality_warnings 2024-02-19 10:49:43.131 7 ERROR oslo_service.service caps = self._host.get_capabilities() 2024-02-19 10:49:43.131 7 ERROR oslo_service.service File "/usr/lib/python2.7/site-packages/nova/virt/libvirt/host.py", line 705, in get_capabilities 2024-02-19 10:49:43.131 7 ERROR oslo_service.service xmlstr = self.get_connection().getCapabilities() 2024-02-19 10:49:43.131 7 ERROR oslo_service.service File "/usr/lib/python2.7/site-packages/nova/virt/libvirt/host.py", line 516, in get_connection 2024-02-19 10:49:43.131 7 ERROR oslo_service.service raise exception.HypervisorUnavailable(host=CONF.host) 2024-02-19 10:49:43.131 7 ERROR oslo_service.service HypervisorUnavailable: Connection to the hypervisor is broken on host: node001 2024-02-19 10:49:43.131 7 ERROR oslo_service.service Regards Richard.

2 1

Removal of boto from Debian
by Thomas Goirand 19 Feb '24

19 Feb '24

Hi, There's currently an ongoing effort to remove boto from Debian. The reason is: there's boto3, and boto is not maintained anymore. These a number of packages that are still using it. Affected packages in the OpenStack world are (at least for Bobcat, we'll see for Caracal): - glance-store - swift - heat-cfntools This IMO requires immediate action from the concerned team, to hopefully get rid of this requirement before Caracal is released. My excuses in advance if the work has already been done. Cheers, Thomas Goirand (zigo)

2 1

[kolla-ansible][Antelope][Senlin] Stop In Progress Updates
by Devin Patterson 19 Feb '24

19 Feb '24

Hello, We are using Senlin to create scaling groups and monitor VM health as well as do rolling updates. Does anyone know of a way to stop an in progress rolling update? For example, if we trigger Senlin to use a new profile, can we stop the update once it is started? Even better, is there a way to connect Senlin health policies to the rolling updates to stop the update if previous node health fails? Regards, Devin

1 0

[neutron] bug deputy report for week of 2024-02-12
by Bence Romsics 19 Feb '24

19 Feb '24

Hi All, We had a handful of new bug reports last week: gate-failure: * https://bugs.launchpad.net/neutron/+bug/2052915 "neutron-ovs-grenade-multinode" and "neutron-ovn-grenade-multinode" failing in 2023.1 and Zed proposed fixes: https://review.opendev.org/q/topic:bug/2052915 * https://bugs.launchpad.net/neutron/+bug/2053227 [ovn-octavia-provider] ovn-octavia-provider-functional-master is broken by ovn build failure proposed fix: https://review.opendev.org/c/openstack/ovn-octavia-provider/+/909074 unassigned: * https://bugs.launchpad.net/neutron/+bug/2053274 [ovn] mtu for metadata veth interface is not set high: * https://bugs.launchpad.net/neutron/+bug/2052937 Policy: binding operations are prohibited for service role proposed fixes: https://review.opendev.org/q/topic:bug/2052937 * https://bugs.launchpad.net/neutron/+bug/2054203 Default flag change in netaddr breaks (at least) unit tests proposed fix: https://review.opendev.org/c/openstack/neutron/+/909317 rfe: * https://bugs.launchpad.net/neutron/+bug/2052906 [RFE] OpenVSwitch HTB QoS not efficient for a lot of interfaces Cheers, Bence

1 0

[Cinder] PTL non-candidacy
by Rajat Dhasmana 18 Feb '24

18 Feb '24

Hello everyone, I'm sending this email out to let everyone know that I will not be running for PTL for the Cinder project during the 2024.2 (Dalmatian) cycle. I started my PTL journey with the Zed cycle and it's been a great experience leading the Cinder project for 4 cycles. I think it is time for someone else to step up and bring their insights into the PTL tasks since I believe a new perspective will surely benefit the project. I would like to thank the Cinder team for their support, without which, it would have been a very hard task to accomplish. Last but not the least, Brian and Jay (former Cinder PTLs) helped out a lot to run things smoothly and in making complex decisions so a special thanks for all their help and support. Thanks Rajat Dhasmana

7 6

[all] Upcoming incompatible changes in netaddr 1.0.x
by Takashi Kajinami 17 Feb '24

17 Feb '24

Hello, While looking into some test failures in an attempt to bump netaddr from 0.10.x to 1.0.x[1], we learned that there are a few changes in netaddr 10.0.x impacting our current usage. I'll attempt to summarize a few important ones in this email, and would like to ask each team to look into your project to assess the impacts caused by these changes. 1. INET_PTON (more restrictive check by default) is used by default In 1.0.1, all features dependent on ip address parsing uses INET_PTON by default instead of INET_ATON. With this change a few "strange" address values such as '192.0' or even '1' is no longer accepted. In case you want to restore the previous behavior then you have to add flags=netaddr.core.INET_ATON to each call. See [2] for example. This is caught in neutron unit tests[4] but my need to be further check in libraries handling IP addresses. 2. valid_ipvN raises exception for non-string value. Previously netaddr.valid_ipv(4|6) returned False for non-string values such as None. This was (unintentionally) changed in 1.0.x. I proposed restoring the old behavior but the maintainer preferred strictness about the input type, and we eventually agreed on throwing more explicit exception (TypeError). At this moment this affects only manila unit tests which I'm fixing by [5], but I wonder if there is any specific use case in any project dependent on the previous behavior. If so then probably we can add our own logic to handle non-string value, for example in oslo_utils.netutils. 3. A few deprecated interfaces are removed. At this moment I'm aware of only impact on heat (caused by implicit_prefix removed), but double-check my be better. Thank you, Takashi [1] https://review.opendev.org/c/openstack/requirements/+/909300 [2] https://review.opendev.org/c/openstack/oslo.utils/+/909307 [3] https://github.com/netaddr/netaddr/issues/368 [4] https://bugs.launchpad.net/neutron/+bug/2054203 [5] https://review.opendev.org/c/openstack/manila/+/909298 -- Takashi Kajinami irc: tkajinam github: https://github.com/kajinamit launchpad: https://launchpad.net/~kajinamit

1 0

[kolla][rabbitmq]: Error when ensuring rabbitmq users exists
by Winicius Allan 17 Feb '24

17 Feb '24

Hello o/, After an upgrade of nodes packages, we have been facing a problem when reconfiguring RabbitMQ and Nova. In the tasks that ensures RabbitMQ users exists the output of task is: FAILED - RETRYING: [xxxx]: nova | Ensure RabbitMQ users exist (5 retries left). FAILED - RETRYING: [xxxx]: nova | Ensure RabbitMQ users exist (4 retries left). FAILED - RETRYING: [xxxx]: nova | Ensure RabbitMQ users exist (3 retries left). FAILED - RETRYING: [xxxx]: nova | Ensure RabbitMQ users exist (2 retries left). FAILED - RETRYING: [xxxx]: nova | Ensure RabbitMQ users exist (1 retries left). I've already tried to redeploy rabbit and reconfigure after that, but does not work neither Anyone can help? I'm unable to perform any operation in my environment. Thanks! Winicius.

3 6