September 2023 - openstack-discuss

[all][debian] First VM running under Bobcat pings: Debian Bobcat repository is ready!
by Thomas Goirand 29 Sep '23

29 Sep '23

Hi, As on each release, my first test is to pop a new OpenStack cluster on my visualized environment (a single machine running 38 VMs with nested virt). Today, my first VM running under Bobcat was ACTIVE and pinged. All of this is available the usual way, especially with extrepo if you want to use the unofficial Bookworm backport: apt-get install extrepo extrepo enable openstack_bobcat apt-get update The following days, I'll be uploading all of this form Debian Experimental to Unstable, trying to add as many SQLAlchemy 2.x support as I can, since the package needs to be uploaded to Unstable soon. There's also a few packages that I didn't have time to finish yet (like the last version of Horizon that has many unit test failures that I didn't have time to look on...). Thanks to everyone that contributed to Bobcat, or that was involved in this release, Cheers, Thomas Goirand (zigo)

2 1

Masakari Issue
by Karera Tony 29 Sep '23

29 Sep '23

Hello Team,, I have a cluster of 4 Compute nodes but when shutdown the VMs so that I can upgrade the kernel for the Hosts, The compute nodes in instance-ha and stuck in Maintenance mode. When I try to change them, I get the error below. Error: Failed to update host. Details ConflictException: 409: Client Error for url: http://10.10.13.31:15868/v1/segments/8d042245-5610-4b84-b611-b633f8f8367c/h…, Host 066c5654-dd1a-4fa9-a664-dad10b89e202 can't be updated as it is in-use to process notifications Regards Tony Karera

1 0

Masakari Issue
by Karera Tony 28 Sep '23

28 Sep '23

Hell oTeam, I have an openstack Wallaby Environment deployed using kolla-ansible. So, I shutdown all the VMs on the compute nodes becauseI wanted to upgrade the nodes kernel. After I rebooted them and all was fine apart from the fact that all the hosts in instance-ha segment are in maitenance mode. When I try to turn to false. I get the error below. *Error: *Failed to update host. Details <http://10.10.13.31/masakaridashboard/hosts/#message_details> ConflictException: 409: Client Error for url: http://x.x.x.x:15868/v1/segments/8d042245-5610-4b84-b611-b633f8f8367c/hosts…, Host 066c5654-dd1a-4fa9-a664-dad10b89e202 can't be updated as it is in-use to process notifications. Regards Tony Karera

1 0

[Glance] Weekly Meeting Cancelled for Today
by Pranali Deore 27 Sep '23

27 Sep '23

Hello Everyone, Cancelling today's weekly meeting, since there is nothing much on agenda for the week & few team members are also not around. See you next week ! Thanks & Regards, Pranali Deore

1 0

[all][tc] python 3.11 testing plan
by Ghanshyam Mann 27 Sep '23

27 Sep '23

Hi All, Some of you are part of discussion for python 3.11 testing but If you are not aware of it, below is the plan for python 3.11 testing in OpenStack. Non voting in 2023.2 ------------------------- You might have seen that python 3.11 job is now running as non voting in all projects[1]. Idea is to run it as non voting for this (2023.2) cycle which will give projects to fix the issue and make it green. As it is running on debian (frickler mentioned the reason of running it in debian in gerrit[2]), it need some changes in bindep.txt file to pass. Here is the example of fix[3] which you can do in your project also. Voting in 2024.1 -------------------- In next cycle (2024.1), I am proposing to make py3.11 testing mandatory [4] and voting (automatically via common python job template). You need to fix the failure in this cycle otherwise it will block the gate once the next cycle development start (basically once 891238 is merged). [1] https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/891227/5 [2] https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/891146/1 [3] https://review.opendev.org/c/openstack/nova/+/891256 [4] https://review.opendev.org/c/openstack/governance/+/891225 [5] https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/891238 -gmann

12 27

[kolla-ansible][neutron] Kolla-ansible deployment where not all network nodes have access to the same external networks
by Moonpiedumplings TheSecond 27 Sep '23

27 Sep '23

I'm currently prototyping for what will be a two node openstack install via kolla-ansible, with one part on a public vps as a network node, and another on my home server as the main compute/controller node, and an allinone node. I want to use the VPS to give virtual machines public ipv6 addresses. However, the VPS only has one ipv4 address (and I am not willing to buy more). I'm thinking the simplest way to have both ipv4 and ipv6 connectivity is just to have my "home" node be a network node as well, and then just virtual machines a ipv6 floating ip, an access to a normal, internal ipv4 subnet. However, I can't get networking working properly, as I don't know how to force openstack to create a network on a specific node. I have it halfway working, I can create a network on the equivalent of the main node in my prototypes, but not on the "vps" node. I can create a network, but not create a virtual machine on that network, or give a virtual machine a floating ip from that network. In my blog, I have documented what I've tried, but it's somewhat disorganized, as I've documented my entire process, not just my networking troubles: https://moonpiedumplings.github.io/projects/build-server-2/#asking-for-help Thanks in advance for any help you have to offer,

1 0

Cinder Bug Report 2023-09-27
by Jon Bernard 27 Sep '23

27 Sep '23

Hello Argonauts, Cinder Bug Meeting Etherpad <https://etherpad.opendev.org/p/cinder-bug-squad-meeting> Undecided - cinder reimage failure leaves the volume in downloading state <https://bugs.launchpad.net/cinder/+bug/2036994> - Status: In Progress - Backup using Swift driver fails when TLS is enabled on internal network <https://bugs.launchpad.net/cinder/+bug/2037113> - Status: New Thanks, -- Jon

1 0

[kolla-ansible] Deploying an kolla-ansible all-in-one to test enforce_scope
by Kyle Burall 27 Sep '23

27 Sep '23

I'm trying to test enabling enforce_scope on a kolla-ansible deployment in prep for rolling it out to an actual deployment and I am having issues with services registering themselves. The playbook fails at registering nova services and the keytsone log says the nova user doesn't exist, as well as the service project and the admin role. I've added enforce_scope and enforce_new_defaults to global.conf, which causes the referenced error. I assume that I'm missing something in the configuration that is required beyond just setting those two oslo_policy options, but I haven't been able to find anything beyond these two settings. Can anyone let me know if there's something specific that I'm missing? Thanks! This email and its attachments, if any, may contain information that is proprietary, business-confidential, and/or legally protected. If you received this communication in error, please notify me by reply message and delete the email and any attachments without copying, distributing, or otherwise using.

1 0

[openstack][nova][cinder] how can we know instance disk usage
by Nguyễn Hữu Khôi 27 Sep '23

27 Sep '23

Hello guys. I am try to get instance disk usage but I cannot. I try to use virsh domblkinfo instance-00000df8 vda but it did not show exactly disk allocation How we get this metric.I used SAN and Ceph as backend storages. Thank you so much, . Nguyen Huu Khoi

1 0

[magnum] Weekly meeting today (27th Sep 2023) cancelled
by Michał Nasiadka 27 Sep '23

27 Sep '23

Hello, Since key members are absent - let’s cancel this weeks (27th Sep 2023) meeting. Best regards, Michal

1 0