December 2024 - openstack-discuss

OSA adding compute without deploy host
by Derek O keeffe 04 Dec '24

04 Dec '24

Hi all, I have a controller and 2 compute setup that was deployed some time ago with OSA. It's been running fine since it was deployed but we now have the need for more compute and I was looking at the deploy host to add a compute host to the cluster. Unfortunately it seems the deploy host was repurposed and I have lost all the config. So two questions, 1) Is it possible to manually add a compute to the cluster (I fairly confident this is a yes) and 2) if so is there any documentation or how to that someone could point me to/share? any help would be greatly appreciated. Regards,Derek

2 2

[kolla-ansible][neutron][openvswitch]
by Adam Tomas 04 Dec '24

04 Dec '24

Hi everyone. I have strange problem with neutron ml2 plugin. I have a little strange setup: controllers in networkA/subnetA computesA in networkA/subnetA computesB in networkB/subnetB (separate L2, controllers can be connected via routing) Because computesB are on different LAN I had to put on two computes in that lan neutron-dhcp-agents and add them to networks served for instances in locationB. In addition aggregates are created so, that if an instance has network from locationB then it has to start on computes in locationB. computes and controllers have one flat network (vlan 21) ans three vlan networks (221, 51 and 251),, vlans 21 and 221 are in locationA, 51 and 251 in locationB, the ml2_plugin is configured like below: [ml2_type_vlan] network_vlan_ranges = physnet1:221:221,physnet2:51:51,physnet3:251:251 bond0 is the neutron_external_interface and has vlan 21 as native and 221,51 and 251 as tagged vlans. Openstack networks are mapped to corresponding physnets. dhcp_agents in location A are working without any problems, vm's are started without any problem in vlans 21 and 221 with addresses from all subnets from networks in these vlans. In location B when I add dhcp_agent the port is doen (namespaces qdhcpxxxxxxxxx are created on computesB and are replying to their own address, but not the gateway or any other address), port status of dhcp_agent is DOWN, and in logs I have: Device Port(admin_state_up=True,allowed_address_pairs=[],binding_levels=[],bindings=[PortBinding],created_at=2024-12-01T17:52:25Z,data_plane_status=None,description='',device_id='dhcpfd2f8da2-397b-5456-a75c-f9c5bbccb8b1-81b9ac42-6267-48e2-8352-0ae5c09d85fa',device_owner='network:dhcp',device_profile=<?>,dhcp_options=[],distributed_bindings=[],dns=None,fixed_ips=[IPAllocation,IPAllocation],hints=None,id=08c60b9d-ea8f-4a2d-a3b7-5f787d15f45d,mac_address=fa:16:3e:37:27:0c,name='',network_id=81b9ac42-6267-48e2-8352-0ae5c09d85fa,numa_affinity_policy=<?>,project_id='c0e2f3d69c064ae5b5596fd150d550d7',qos_network_policy_id=None,qos_policy_id=None,revision_number=2,security=PortSecurity(08c60b9d-ea8f-4a2d-a3b7-5f787d15f45d),security_group_ids=set([]),status='DOWN',updated_at=2024-12-01T17:52:25Z) is not bound. Device 08c60b9d\-ea8f\-4a2d\-a3b7\-5f787d15f45d not defined on plugin or binding failed Port 08c60b9d-ea8f-4a2d-a3b7-5f787d15f45d cannot update to ACTIVE because it failed. _port_provisioned /var/lib/kolla/venv/lib/python3.10/site-packages/neutron/plugins/ml2/plugin.py:357 Why the same configuration works in LocationA? In location B the same situation is with all ports - even if I create a port "by hand". The error message (in debug mode!) says not much - Device 08c60b9d\-ea8f\-4a2d\-a3b7\-5f787d15f45d not defined on plugin or binding failed, how can I find what is wrong? I would appreciate any help. Best regards, Adam Tomas

1 1

[kolla-ansible][nova] Refresh cell cache in nova scheduler
by Satish Patel 04 Dec '24

04 Dec '24

Folks, What is wrong here, I do have the following hostname resolvable from all the nodes even if they are in the /etc/hosts file, Why is the following task saying UNREACHABLE and from where its trying to reach ? PLAY [Refresh nova scheduler cell cache] ***************************************************************************************************************************************************** TASK [nova : Refresh cell cache in nova scheduler] ******************************************************************************************************************************************* fatal: [os2-bos-ctrl01]: UNREACHABLE! => {"changed": false, "msg": "Data could not be sent to remote host \"os2-bos-ctrl01\". Make sure this host can be reached over ssh: ", "unreachable": true} fatal: [os2-bos-ctrl02]: UNREACHABLE! => {"changed": false, "msg": "Data could not be sent to remote host \"os2-bos-ctrl02\". Make sure this host can be reached over ssh: ", "unreachable": true} fatal: [os2-bos-ctrl03]: UNREACHABLE! => {"changed": false, "msg": "Data could not be sent to remote host \"os2-bos-ctrl03\". Make sure this host can be reached over ssh: ", "unreachable": true}

4 5

[octavia] [lbaas] [ovs] Is octavia sr-iov supporting tunnels?
by yardalgedal＠gmail.com 04 Dec '24

04 Dec '24

Hi, Openstack Team, Is new octavia sr-iov feature supporting vxlan tunnels like in ovs or it can work only with provider networks (vlan)?

3 2

[kayobe] Running seed on ansible control host
by Olaf Trygve Berglihn 04 Dec '24

04 Dec '24

Dear list, Can someone explain how to configure kayobe to set up the seed hypervisor and seed vm on the ansible control host (where you run 'kayobe seed hypervisor host configure' and 'kayobe seed vm provision')? Thanks! -Olaf

2 5

[masakari] failures with new version of Eventlet
by Herve Beraud 04 Dec '24

04 Dec '24

Hey Masakari maintainers, Masakari is failing with new the version of Eventlet https://review.opendev.org/c/openstack/requirements/+/933257 This job failure indicates that Masakari will surely stop working correctly once this version bump will be merged, and we need that requirements bump to support newer versions of Python in OpenStack, so... Masakari is a non-voting job, so even if this job is failing, the requirement bump will proceed, but that would be better if every project is compatible and ready to go. If you need help do not hesitate to: - join us on #openstack-eventlet-removal; - join us at our next meeting (Dec 17) https://wiki.openstack.org/wiki/Eventlet-removal#Recurring_Events. Thanks for your hard work. -- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/

1 0

[horizon] Upstream weekly meeting cancelled Dec 4
by Tatiana Ovchinnikova 03 Dec '24

03 Dec '24

Hi team, Since half of our team is on PTO tomorrow, I am cancelling the meeting. See you all on December 11! With kind regards, Tatiana

1 0

[OSSA-2024-005] Neutron: Authorization bypassed when setting tags on Neutron networks (CVE-2024-53916)
by Jay Faulkner 03 Dec '24

03 Dec '24

=========================================================================== OSSA-2024-005: Authorization bypassed when setting tags on Neutron networks =========================================================================== :Date: December 03, 2024 :CVE: CVE-2024-53916 Affects ~~~~~~~ - Neutron: >=23.0.0 <23.2.1, >=24.0.0 <24.0.2, >=25.0.0 <25.0.1 Description ~~~~~~~~~~~ Tore Anderson of Redpill Linpro AS discovered that Neutron does not apply the proper policy check for changing network tags. An unprivileged tenant is able to change (add and clear) tags on network objects which do not belong to the tenant, and this action is not being subjected to the proper policy authorization check. Patches ~~~~~~~ -https://review.opendev.org/c/openstack/neutron/+/936849 (2023.2/bobcat) -https://review.opendev.org/c/openstack/neutron/+/936846 (2024.1/caracal) -https://review.opendev.org/c/openstack/neutron/+/936843 (2024.2/dalmatian) -https://review.opendev.org/c/openstack/neutron/+/935883 (2025.1/epoxy) Credits ~~~~~~~ - Tore Anderson from Redpill Linpro AS (C, V, E, -, 2, 0, 2, 4, -, 5, 3, 9, 1, 6) References ~~~~~~~~~~ -https://launchpad.net/bugs/2088986 -http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2024-53916 -- Jay Faulkner OpenStack VMT

1 0

[publiccloud-sig] Reminder: next meeting December 4th (tomorrow) - 0800 UTC
by Felix Kronlage-Dammers 03 Dec '24

03 Dec '24

Hi everyone, This Wednesday, tomorrow, (December 4th) the next meeting of the publiccloud-sig is going to happen. We’ll be meeting at 0800 UTC here: https://meetpad.opendev.org/publiccloud-sig <https://meetpad.opendev.org/publiccloud-sig> See also here for all other details: https://wiki.openstack.org/wiki/PublicCloudSIG <https://wiki.openstack.org/wiki/PublicCloudSIG> See you tomorrow! felix -- Felix Kronlage-Dammers 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

[nova] Rescuing a volume-backed instance
by Francesco Di Nucci 03 Dec '24

03 Dec '24

Hello, I am using Openstack 2024.1 and I have issues with an instance with a corrupted FS - I wanted to start it in rescue mode, but I'm getting stuck After reading nova#1926601 <https://bugs.launchpad.net/nova/+bug/1926601> , downloaded System Rescue CD and created an image with properties [otherwise it always failed with "Cannot rescue a volume-backed instance (HTTP 400)"] : openstack image create --file systemrescue-11.02-amd64.iso --disk-format iso --container-format bare --property hw_rescue_bus=virtio --property hw_rescue_device=disk systemrescuecd Then tried to rescue the instance: openstack --os-compute-api-version 2.96 server rescue --image systemrescuecd (my-instance-id) Now it fails with "Instance (my-instance-id) cannot be rescued: Driver Error: Image (my-image-id) is unacceptable: Image not in a supported format" How should I proceed? Thanks in advance -- Francesco Di Nucci System Administrator Compute & Networking Service, INFN Naples Email: francesco.dinucci(a)na.infn.it

3 3