June 2024 - openstack-discuss - lists.openstack.org

[tc][all] OpenStack Technical Committee Weekly Summary and Meeting Agenda (2024.2/R-16)
by Goutham Pacha Ravi 11 Jun '24

11 Jun '24

Hello Stackers, We're three weeks away from Milestone-2 of the 2024.2 (Dalmatian) release cycle [1]. Teams are pursuing mid-cycle meetings and design specification freezes in the next couple of weeks. I'd like to begin this week's recap with a shoutout to the TaCT SIG and especially Dr. Jens Harbott (frickler) for helping us with our PyPi cleanup efforts [2]. Apart from this; we had very few governance changes in the past week. I want to take the opportunity to remind you about our gate hygiene efforts. As a contributor, if the continuous integration system (Zuul) is upset with your change, please avoid re-testing with a simple "recheck" comment (or even a comment like "recheck, unrelated failure"). We're trying to isolate unstable CI tests, and if you flag specific failing tests or infrastructure issues, we might be able to address them better. === Weekly Meeting === Last week's meeting was short, and was held simultaneously on Video [3] and IRC [4]. We discussed adding a second maintainer account on all OpenStack packages on PyPi that would be able to push package updates and make project metadata changes (such as adding/removing maintainers). We'll be incorporating this new account soon. We also spoke about graduating Skyline from the list of "emerging projects" [5]. The project team guide has a lot of criteria to create and maintain projects in the OpenStack way; these criteria however, are applied subjectively and are flexible to accommodate specific situations, such as not requiring weekly meetings or formal presence at the PTG. These activities are encouraged. Skyline's contributors are small in number, but have welcomed new contributors. Our recent user surveys have shown that deployments have started adopting Skyline; and installer/configuration management services are seeing feature requests to include Skyline. So the TC feels we need to take this project off of the "emerging" projects list. We volunteered Dr. Jens (frickler) to assess the situation and coordinate this. Our next weekly meeting is tomorrow (June 11th 2024) at 1800 UTC, exclusively on OFTC's #openstack-tc channel. Please see the agenda at [6]. Remember, you can always propose topics to this agenda; and if you're trying to bring up something on short notice, I'll make the attempt to leave some time aside for Open Discussion. === Governance Proposals === ==== Merged ==== Appoint ChunYang Wu as PTL of Trove | https://review.opendev.org/c/openstack/governance/+/920462 ==== Open for review/discussion ==== Remove Eventlet From Openstack | https://review.opendev.org/c/openstack/governance/+/902585 Add TC liaison in DPL model implementation | https://review.opendev.org/c/openstack/governance/+/917516 Update criteria for the inactive projects to become active again | https://review.opendev.org/c/openstack/governance/+/921500 Update to include docs and miscellaneuos repos for AC status | https://review.opendev.org/c/openstack/governance/+/915021 Thank you very much for reading! On behalf of the OpenStack TC, Goutham Pacha Ravi (gouthamr) OpenStack TC Chair [1] https://releases.openstack.org/dalmatian/schedule.html [2] https://etherpad.opendev.org/p/openstack-pypi-maintainers-cleanup#L52 [3] https://www.youtube.com/watch?v=34w1Aoai1EA [4] https://meetings.opendev.org/meetings/tc/2024/tc.2024-06-04-18.00.log.html [5] https://governance.openstack.org/tc/reference/emerging-technology-and-inact… [6] https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Next_Meeting

1 0

[manila] stuck at connecting ceph backend with manila
by Shubham Kumar Yadav 11 Jun '24

11 Jun '24

hello, I'm trying to install manila in my openstack cloud but having problem with "*Authorizing the driver to communicate with Ceph*" *1 -> *i have created a configmap with these confs and mounted it on manila-share pod [client.manila]client mount uid = 0client mount gid = 0log file = /opt/stack/logs/ceph-client.manila.logadmin socket = /opt/stack/status/stack/ceph-$name.$pid.asokkeyring = /etc/ceph/manila.keyring *2 -> *then i have created a secret with client.manila keyring and attached it to the manila-share pod at keyring = /etc/ceph/manila.keyring *3 -> *inside manila-share pod i see the logs 2024-06-03 09:00:53.682 6 INFO manila.share.drivers.cephfs.driver [None req-22cd7b29-a383-44cd-8cb8-1c639a3034dc - - - - - -] [CEPHFSNATIVE1] Ceph client found, connecting... 2024-06-03 09:00:53.713 6 ERROR manila.share.manager [None req-22cd7b29-a383-44cd-8cb8-1c639a3034dc - - - - - -] Error encountered during initialization of driver CephFSDriver@lab-11053-26006-ceph-2@cephfsnative1: manila.exception.ShareBackendException: [%(be)s] Ceph client failed to connect. 2024-06-03 09:00:53.713 6 ERROR manila.share.manager Traceback (most recent call last): 2024-06-03 09:00:53.713 6 ERROR manila.share.manager File "/var/lib/openstack/lib/python3.10/site-packages/manila/share/drivers/cephfs/driver.py", line 421, in rados_client 2024-06-03 09:00:53.713 6 ERROR manila.share.manager self._rados_client.connect() 2024-06-03 09:00:53.713 6 ERROR manila.share.manager File "rados.pyx", line 690, in rados.Rados.connect 2024-06-03 09:00:53.713 6 ERROR manila.share.manager rados.IOError: [errno 5] RADOS I/O error (error connecting to the cluster) 2024-06-03 09:00:53.713 6 ERROR manila.share.manager 2024-06-03 09:00:53.713 6 ERROR manila.share.manager During handling of the above exception, another exception occurred: 2024-06-03 09:00:53.713 6 ERROR manila.share.manager 2024-06-03 09:00:53.713 6 ERROR manila.share.manager Traceback (most recent call last): 2024-06-03 09:00:53.713 6 ERROR manila.share.manager File "/var/lib/openstack/lib/python3.10/site-packages/manila/share/manager.py", line 371, in _driver_setup 2024-06-03 09:00:53.713 6 ERROR manila.share.manager self.driver.do_setup(ctxt) 2024-06-03 09:00:53.713 6 ERROR manila.share.manager File "/var/lib/openstack/lib/python3.10/site-packages/manila/share/drivers/cephfs/driver.py", line 274, in do_setup 2024-06-03 09:00:53.713 6 ERROR manila.share.manager rados_client=self.rados_client, 2024-06-03 09:00:53.713 6 ERROR manila.share.manager File "/var/lib/openstack/lib/python3.10/site-packages/manila/share/drivers/cephfs/driver.py", line 424, in rados_client 2024-06-03 09:00:53.713 6 ERROR manila.share.manager raise exception.ShareBackendException( 2024-06-03 09:00:53.713 6 ERROR manila.share.manager manila.exception.ShareBackendException: [%(be)s] Ceph client failed to connect. 2024-06-03 09:00:53.713 6 ERROR manila.share.manager 2024-06-03 09:00:53.714 6 DEBUG manila.utils [None req-22cd7b29-a383-44cd-8cb8-1c639a3034dc - - - - - -] Finished call to 'manila.share.manager.ShareManager.init_host.<locals>._driver_setup' after 2222.651(s), this was the 12th time calling it. log_it /var/lib/openstack/lib/python3.10/site-packages/tenacity/after.py:44 2024-06-03 09:00:53.714 6 DEBUG manila.utils [None req-22cd7b29-a383-44cd-8cb8-1c639a3034dc - - - - - -] Retrying manila.share.manager.ShareManager.init_host.<locals>._driver_setup in 600.0 seconds as it raised ShareBackendException: [%(be)s] Ceph client failed to connect.. log_it /var/lib/openstack/lib/python3.10/site-packages/tenacity/before_sleep.py:65 *4 -> *My ceph cluster is healthy 5 -> My manila confs look like (what could be wrong here)? conf: manila: DEFAULT: debug: true enabled_share_protocols: NFS,CIFS,CEPHFS enabled_share_backends: cephfsnative1 cephfsnative1: driver_handles_share_servers: False share_backend_name: CEPHFSNATIVE1 share_driver: manila.share.drivers.cephfs.driver.CephFSDriver cephfs_conf_path: /etc/ceph/ceph.conf cephfs_protocol_helper_type: CEPHFS cephfs_auth_id: manila cephfs_cluster_name: rook-ceph ( my rook-ceph cluster name) cephfs_filesystem_name: myfs (i created this fs manually) *6 -> P*ermissions for the user manila ceph auth get-or-create client.manila -o manila.keyring \ mgr 'allow rw' \ mon 'allow r' Hope to hear from you soon with a solution Thank you!

2 1

[Neutron] Bug deputy report for week of June 3rd
by Jakub Libosvar 10 Jun '24

10 Jun '24

Hi, here is the bug deputy report. There are two bugs that require attention: - One seems to be an old issue with conntrack sync, I put it to the wishlist but maybe can be closed as a duplicate. - Another one doesn’t contain enough information to be triaged, please see below. Kuba Needs attention --------------- * Packets dropped in br-int bridge - https://bugs.launchpad.net/neutron/+bug/2067851 - It seems like report needs help troubleshooting the problem or use some accelerated datapath like DPDK. High ---- * Issue associating floating IP with OVN load balancer - https://bugs.launchpad.net/neutron/+bug/2068644 - Patch proposed by Will Szumski: https://review.opendev.org/c/openstack/neutron/+/921663 * SFC: networking-sfc-tempest failures - https://bugs.launchpad.net/neutron/+bug/2068727 - Patch proposed by Lajos: https://review.opendev.org/c/openstack/networking-sfc/+/921514 * Using "joined" relation on tags can cause slow query times in certain situations - https://bugs.launchpad.net/neutron/+bug/2068761 - Revert proposed by Miro: https://review.opendev.org/c/openstack/neutron/+/921566 Low --- * Neutron should have better regex matching in zuul jobs - https://bugs.launchpad.net/neutron/+bug/2068523 - Assigned to Brian Wishlist -------- * [ovn] Dose L3 router ha support sync conntrack states - https://bugs.launchpad.net/neutron/+bug/2068704 - This was discussed at the last PTG in April and probably can be closed as a duplicate of https://bugs.launchpad.net/neutron/+bug/1365438 Invalid ------- * No support on using domain name for ovn connection string in ml2_conf.ini - https://bugs.launchpad.net/neutron/+bug/2067871 - It’s been pointed out on the bug that the issue is fixed in OVS

1 0

[ironic] Notes from Baremetal SIG, June 5 2024
by Jay Faulkner 10 Jun '24

10 Jun '24

Hi all, The Bare Metal SIG met up Wednesday, June 5 2024 at CERN – one day before the large OpenInfra Days event. There were over 2 dozen Ironic operators and contributors present. Riccardo gave a presentation overviewing Metal3. JayF gave a presentation covering the basics of most Ironic features and highlighted several features which were deprecated. While the talks were recorded, I don't believe the recordings are available yet – I do plan on recreating an updated version of my talk for Youtube in the near future for anyone interested. After the presentations and lunch, participants returned to discuss their use cases and ask questions relating to the earlier presentations. The discussion, as well as specific action items/follow-ups, were documented here: https://etherpad.opendev.org/p/baremetal-sig-cern-june-5-2024 – I will be documenting these action items and RFEs as appropriate over the next week. Add in a short afternoon break to tour the CERN data center, seeing the thousands of machines provisioned by Ironic for CERN, and it was a pleasant day discussing and building rapport with the Ironic community. In my over ten years of working in Ironic, this was the most productive operator/contributor feedback session I've had the pleasure of attending – the attendees were engaged and I'm grateful for their participation and attendance. A special thanks to: - Dmitry Tantsur and Julia Kreger connected up remotely to the session to help field questions. Their expertise was a great help! - Arne Wieback and his team at CERN were gracious hosts and provided us a great VIP tour of the datacenter space. If you were at this event, and have feedback on how we can make future ones better, feel free to reach out here or privately with your ideas. Thanks, Jay Faulkner

1 0

[designate] [neutron] DNS and SSL error 524297
by Francesco Di Nucci 10 Jun '24

10 Jun '24

Hi all, I am using Designate with WSGI and switched to a SSL setup after tests. Somehow now there are floating IPs with a DNS name/entry that do not show in DNS zones, and when I try to release them, the neutron WSGI on network node reports this error: /SSL exception connecting to https://my-controller.example.com:9001/zones?name=zone.cloud.example.com.: HTTPSConnectionPool(host='my-controller.example.com', port=9001): Max retries exceeded with url: /zones?name=zone.cloud.example.com. (Caused by SSLError(SSLError(524297, '[SSL] PEM lib (_ssl.c:4065)')))/ I thought of an SSL config error, but the setup is identical to other working endpoints, and when testing with "/openssl s_client -connect my-controller.example.com:9001"/, everything just works Has anyone encountered this error? Thanks in advance Francesco Di Nucci

1 0

[blazar] Next IRC meeting cancelled
by Pierre Riteau 08 Jun '24

08 Jun '24

Hello, Due to unavailability of key participants, the next IRC meeting scheduled for June 13 is cancelled. We will meet again on June 27. Pierre Riteau (priteau)

1 0

[neutron] Please hold off on merging patches
by Brian Haley 07 Jun '24

07 Jun '24

Hi Neutrinos, It was noticed today by Mohammed Naser that the neutron gate was only running the docs and pep8 jobs. We were able to track it down to the RE2 patches that were recently merged [0]. Until the reverts land please don't +W anything since it could merge untested and would just need to be reverted. Other projects might have similar issues if they updated code using the 'negate' directive for regex recently as we found it doesn't actually work with RE2 when using negative lookahead [1]. Thanks, -Brian [0] https://review.opendev.org/q/topic:"files-negate" [1] https://github.com/google/re2/issues/156

4 8

oslopolicy-policy-generator: what is the namespace to use to get Mgnum policy?
by Michel Jouvin 07 Jun '24

07 Jun '24

Hi, I'd like to see the default Magnum (Oslo) policy used with Magnum. With other services, I use oslopolicy-policy-generator and use the service name as the namespace but for Magnum, "--namespace magnum" returns an error saying that there is no "magnum" namespace. Does anybody know what is the namespace to use with Magnum? Or how to display the currently active policy, preferably in YAML format? Thanks in advance. Cheers, Michel

2 6

[swift] clean up orphan segment parts
by Rafael Tavares Oliveira 07 Jun '24

07 Jun '24

Hey there! We've been using swift and we came across this bug https://bugs.launchpad.net/swift/+bug/1813202. We decided to work it out and have an ongoing patch to solve this issue alongside ronaldo.maia(a)luizalabs.com, we would like to share and get some feedback before we submit it. To solve this, we took inspiration in obj/expirer.py: In S3 controllers we register task objects in an internal account, each task object points to an orphaned slo object, which can be created in two situations, when a multipart-upload is not completed or when a multipart object is overwritten. These tasks will be checked by the daemon (a new one we made called "s3-slo-expirer") after some time and deleted from the base accounts. Rafael

2 2

[all][tc] Eventlet migration: Call to action
by Jay Faulkner 07 Jun '24

07 Jun '24

Hi all, As you may have read in Goutham's TC summary, the existing eventlet goal[0] is unlikely to gain consensus to merge. This appears to be due to the difference in needs between OpenStack projects, and belief that it's unlikely a single solution would serve them all. Our current situation is: 1) Eventlet usage must be discontinued. The temporary maintainers are just that -- temporary -- and we should not expect eventlet to continue to be a valid path moving forward. 2) We have basic agreement that shared tooling, such as oslo libraries, should not dictate a specific threading model. In lieu of specific technical leadership and guidance from the TC, this leaves it to individual projects to decide what threading model to use instead of eventlet monkey patching. The unmerged spec[0] may be a good starting point if you're wondering where to start. Importantly: do not let this work wait! It is a significant change and will take a long time to complete. Thanks, Jay Faulkner 0: https://review.opendev.org/c/openstack/governance/+/902585

6 7