May 2025 - openstack-discuss - lists.openstack.org

[barbican] Cancel Weekly Meeting
by Mauricio Harley 22 May '25

22 May '25

Hi, guys! We need to cancel the weekly meeting that would happen on Monday the 26th. This is necessary since I'll be on PTO that day. Looking forward to meeting you all on the Monday following: June the 2nd. Best regards, Mauricio

1 0

#eventlet-removal - Announcing the New Threading Backend of oslo.service
by Herve Beraud 22 May '25

22 May '25

I'm happy to announce the new version of oslo.service is out! [1] This new version came out with the new Threading backend [2]. OpenStack Upper constraints are on the way to be upgraded to this new version [3]. You can start using the new threading backend by following the steps below: ``` import oslo_service.backend as service service.init_backend(service.BackendType.THREADING) ``` Ideally this backend selection should be made in your patcher module. Almost all OpenStack services own a household patcher. Neutron and Nova already demonstrated a successful adoption of this new backend on Nova scheduler and Neutron L3 agent: - https://review.opendev.org/c/openstack/neutron/+/938404 - https://review.opendev.org/c/openstack/nova/+/948450/17 Devstack is on the way to be adapted to allow you to set up `SYSTEMD_ENV_VARS`. [4] This will allow passing environment variables from zuul job definitions to systemd service files via the local conf. During the Eventlet removal effort this will allow teams to have separate jobs testing the same service in different concurrency modes. The first use case of this is to implement scenarios like passing variable `OS_NOVA_DISABLE_EVENTLET_PATCHING=true` to nova services that already supports running in native threading mode instead of with Eventlet. To get a more concrete integration example, you can take inspiration from Nova patches [5][6] to see how to implement your backend switch on your deliverables. A comprehensive guide is currently under creation. [7] Next steps will be to address a couple of remaining comments, to introduce functional tests, and to democratize the usages of this new backend. Thank you to Daniel for his hard work and all his efforts on this topic. Also I wanted to specially thank Gibi and Rodolfo for their close collaborations on this new backend implementation. Please join me to warmfully thank them. Our mission is to equip the community with everything it needs to leave Eventlet behind. This backend is only a tiny piece of this giant puzzle. Your turn to place your piece: https://removal.eventlet.org Cheers, [1] https://review.opendev.org/c/openstack/releases/+/950646 [2] https://review.opendev.org/c/openstack/oslo.service/+/945720 [3] https://review.opendev.org/c/openstack/requirements/+/950653 [4] https://review.opendev.org/c/openstack/devstack/+/948436 [5] https://review.opendev.org/c/openstack/nova/+/948311 [6] https://review.opendev.org/c/openstack/nova/+/948450 [7] https://review.opendev.org/c/openstack/oslo.service/+/940664 -- Hervé Beraud Principal Software Engineer at Red Hat irc: hberaud https://github.com/4383/

1 0

"quantum" package name on pypi
by JDK Phoenix 21 May '25

21 May '25

Dear Openstack-Team, I am the creator and maintainer of the programming language "qb" (qb · PyPI <https://pypi.org/project/qb/>, qb <https://lschreiber.eu/qb/>) and I wanted to implement a package manager similar to how Maven, Cargo or Poetry works. I picked the name "quantum" for this program and noticed that this name was already in use on pypi for a package developed by you. This program has only a singular contribution in 2012 and seems to be inactive since then. I wanted to ask if it would be possible for you to transfer the quantum-package name on pypi to my user (Leander420). I would be very grateful if this could be possible. I thank you for your time and consideration Kind regards Leander S. Schreiber

6 6

Issue on IP Assignment on newly created VM after deleting old one
by sudeb ghosh 21 May '25

21 May '25

Hi Team, We are operating an OpenStack environment managed by Ansible. Openstack Version: Ussuri(21.2.1) OS: CentOS8 We had a volume backed VM named VM01 with IP of 10.10.10.10 for example. We required to delete the VM01 & recreate a VM named VM02 using the same volume. We used the same project & same network what we used for VM01 earlier. VM02 is getting any one available IPs from the given network. Manually i am able to detach the existing IP & attached the specific IP but its not pinging. Query: i) What is the option if i need the same IP(10.10.10.10) will be assigned to VM02 as well autometically? ii) What is fix to get the said old IP(10.10.10.10) working if i do attach manually in GUI through newtwork-interface? BR// Sudeb Ghosh 7044064878 9332034788

3 3

[openstack][nova]qemu system wont release memory
by Nguyễn Hữu Khôi 21 May '25

21 May '25

Hello. I have an OpenStack instance that was using 14GB out of 16GB of memory. Later, when it was using only 4GB, I checked the QEMU process on the compute node, and it was still consuming 14GB, this means memory isn't released when usage decreases. Do we have solutions for this? I use Ubuntu 24.04 instances. Nguyen Huu Khoi

2 2

[watcher] Call for maintainers - MAAS integration
by Douglas Viroel 20 May '25

20 May '25

Hi everyone, At the last watcher's weekly meeting[1], the team discussed the impacts of the eventlet removal in Watcher's code, and it was brought up that the integration with MAAS (Metal As a Service) support is affected. At the moment, we do not have any active members in Watcher that have any experience with MAAS. CI also doesn't have any job running and testing this integration, and we don't have documentation of how it works. This is a call for maintainers interested in supporting MAAS in Watcher to help fill these gaps. It is worth mentioning that, as discussed in the PTG[2], we also plan to mark this integration as *experimental* once it does not provide minimal testing and documentation (this is also the case of other integrations). But we will cover this topic in a different thread. [1] https://meetings.opendev.org/meetings/watcher/2025/watcher.2025-05-15-12.01… [2] https://etherpad.opendev.org/p/apr2025-ptg-watcher#L71 Thanks and regards, -- Douglas Viroel - dviroel

1 0

[ops][neutron] Neutron-db-manage upgrade from Antelope to Bobcat fails with "foreign key constraint 'porthints_ibfk_1' are incompatible" error
by federica fanzago 20 May '25

20 May '25

Dear all, we are running an OpenStack deployment Yoga on AlmaLinux 9.5 and we are attempting the upgrade Openstack up to Caracal release, one release at time (Zed--Antelope--Bobcat--Caracal) The procedure we are following is to shut down all openstack services, to remove the "old" openstack release and to install the new one, upgrading all service databases. This procedure works well for Zed and Antelope, but during the installation of the Bobcat release, the Neutron database upgrade fails when creating the new "porthints" table. The command we are using is su -s /bin/sh -c "neutron-db-manage --config-file /etc/neutron/neutron.conf --config-file /etc/neutron/plugins/ml2/ml2_conf.ini upgrade head" neutron and the error message we receive is ... File "/usr/lib/python3.9/site-packages/pymysql/err.py", line 107, in raise_mysql_exception raise errorclass(errno, errval) sqlalchemy.exc.OperationalError: (pymysql.err.OperationalError) (3780, "Referencing column 'port_id' and referenced column 'id' in foreign key constraint 'porthints_ibfk_1' are incompatible.") [SQL: CREATE TABLE porthints ( port_id VARCHAR(36) NOT NULL, hints VARCHAR(4095) NOT NULL, PRIMARY KEY (port_id), FOREIGN KEY(port_id) REFERENCES ports (id) ON DELETE CASCADE )ENGINE=InnoDB ... Our "ports" table uses the collation utf8mb3_general_ci mysql> SHOW FULL COLUMNS FROM ports; +------------------+--------------+--------------------+------+-----+---------+-------+---------------------------------+---------+ | Field | Type | Collation | Null | Key | Default | Extra | Privileges | Comment | +------------------+--------------+--------------------+------+-----+---------+-------+---------------------------------+---------+ | project_id | varchar(255) | utf8mb3_general_ci | YES | MUL | NULL | | select,insert,update,references | | | id | varchar(36) | utf8mb3_general_ci | NO | PRI | NULL | | select,insert,update,references | | | name | varchar(255) | utf8mb3_general_ci | YES | | NULL | | select,insert,update,references | | | network_id | varchar(36) | utf8mb3_general_ci | NO | MUL | NULL | | select,insert,update,references | | | mac_address | varchar(32) | utf8mb3_general_ci | NO | | NULL | | select,insert,update,references | | | admin_state_up | tinyint(1) | NULL | NO | | NULL | | select,insert,update,references | | | status | varchar(16) | utf8mb3_general_ci | NO | | NULL | | select,insert,update,references | | | device_id | varchar(255) | utf8mb3_general_ci | NO | MUL | NULL | | select,insert,update,references | | | device_owner | varchar(255) | utf8mb3_general_ci | NO | | NULL | | select,insert,update,references | | | standard_attr_id | bigint | NULL | NO | UNI | NULL | | select,insert,update,references | | | ip_allocation | varchar(16) | utf8mb3_general_ci | YES | | NULL | | select,insert,update,references | | +------------------+--------------+--------------------+------+-----+---------+-------+---------------------------------+---------+ in Bobcat the collation required for the new "porthints" table is latin1_swedish_ci MariaDB [neutron]> show create table porthints; +-----------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+ | Table | Create Table | +-----------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+ | porthints | CREATE TABLE `porthints` ( `port_id` varchar(36) NOT NULL, `hints` varchar(4095) NOT NULL, PRIMARY KEY (`port_id`), CONSTRAINT `porthints_ibfk_1` FOREIGN KEY (`port_id`) REFERENCES `ports` (`id`) ON DELETE CASCADE ) ENGINE=InnoDB DEFAULT CHARSET=latin1 COLLATE=latin1_swedish_ci | So, we have tried to create the porthints table in mysql, forcing the collation to utf8mb3_general_ci mysql> CREATE TABLE porthints ( port_id VARCHAR(36) NOT NULL, hints VARCHAR(4095) NOT NULL, PRIMARY KEY (port_id), FOREIGN KEY(port_id) REFERENCES ports (id) ON DELETE CASCADE )ENGINE=InnoDB DEFAULT CHARSET=utf8 DEFAULT COLLATE utf8mb3_general_ci; and commenting out the creation of the new table in the neutron code /usr/lib/python3.9/site-packages/neutron/db/migration/alembic_migrations/versions/2023.2/expand/6f1145bff34c_port_hints.py. In this way, the upgrade works, but we encounter the same kind of problem during the upgrade to Caracal, specifically with the creation of the new table porthardwareoffloadtype. Have you already encountered this issue during the Neutron upgrade? Do you have any suggestions on how to upgrade the neutron DB, without manually creating new tables? Thanks a lot, cheers, Federica -- Federica Fanzago INFN Sezione di Padova Via Marzolo, 8 35131 Padova - Italy Tel: +39 049.967.7367 --

4 8

Osprofiler and nova
by engineer2024 20 May '25

20 May '25

Team, I am facing an issue with osprofiler in nova, where the 'trace show' is failing. I have filed a bug at https://bugs.launchpad.net/osprofiler/+bug/2111345 But I don't see any recent activity there. Is the osprofiler project active ? Please guide me as per the issue mentioned there... Thanks elinux

1 0

[publiccloud-sig] Next meeting of Public Cloud SIG - May 21st at 0700 UTC
by Felix Kronlage-Dammers 20 May '25

20 May '25

Hi everyone, This upcoming Wednesday (May 21st - this is tomorrow :) the next meeting of the publiccloud-sig is going to happen. We’ll be meeting at 0700 UTC here: https://meetpad.opendev.org/publiccloud-sig See also here for all other details: https://wiki.openstack.org/wiki/PublicCloudSIG See you tomorrow! cheers felix -- Felix Kronlage-Dammers 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

CEPH deployment method: cephadm vs puppet
by Francesco Di Nucci 20 May '25

20 May '25

Hi all, we're planning (finally) to setup a CEPH cluster to be used as OpenStack backend. The cloud is actually set up with Foreman+Puppet, to setup CEPH what would you advice? cephadm as it's the preferred method in the CEPH docs or Puppet with the puppet-ceph module, as it's part of openstack? Thanks in advance -- Francesco Di Nucci System Administrator Compute & Networking Service, INFN Naples Email: francesco.dinucci(a)na.infn.it

6 14