May 2023 - openstack-discuss - lists.openstack.org

[neutron] Bug Deputy Report May 15 - May 21
by Elvira Garcia Ruiz 22 May '23

22 May '23

Hi everyone! I was the bug deputy last week. There are two critical bugs reported, both have fixes ready for review. Find the full report below: Critical --------- - https://bugs.launchpad.net/neutron/+bug/2019802/ - [master] Not all fullstack tests running in CI Fix proposed by yatin: https://review.opendev.org/c/openstack/neutron/+/883120 - https://bugs.launchpad.net/neutron/+bug/2019946 - [S-RBAC] context.elevated() method from neutron-lib should ensure all required roles are set in context object Fix: https://review.opendev.org/c/openstack/neutron-lib/+/883345 Assigned to Slawek High ------ https://bugs.launchpad.net/neutron/+bug/2020050 - [sqlalchemy-20] Use a ``TextClause`` object for the empty strings in the DB model definitions Assigned to Rodolfo https://bugs.launchpad.net/neutron/+bug/2020195 - [ovn-octavia-provider] functional test intermittently fail with DB error: Cursor needed to be reset because of commit/rollback and can no longer be fetched from. Related commit: https://review.opendev.org/c/openstack/ovn-octavia-provider/+/883662 Assigned to Fernando - https://bugs.launchpad.net/neutron/+bug/2020215 - ml2/ovn refuses to bind port due to dead agent randomly in the nova-live-migrate ci job Fix: https://review.opendev.org/c/openstack/neutron/+/883687 Assigned to Sean Mooney Medium ----------- - https://bugs.launchpad.net/neutron/+bug/2019887 - [OVN] Log object deletion fails if related security group was simultaneously modified Fix: https://review.opendev.org/c/openstack/neutron/+/883102 Assigned to Elvira - https://bugs.launchpad.net/neutron/+bug/2019859 - stable/xena regression - NotImplementedError: Operator 'getitem' is not supported on this expression Fix: https://review.opendev.org/c/openstack/neutron/+/883288 Assigned to Rodolfo - https://bugs.launchpad.net/neutron/+bug/2019948 - [alembic] Alembic operations require keywords only arguments Fix: https://review.opendev.org/c/openstack/neutron/+/883340 Assigned to Rodolfo - https://bugs.launchpad.net/neutron/+bug/2020058 - [OVN][OVN-BGP-AGENT] Expose port hosting information for virtual ports In Progress Assigned to Lucas Alvares - https://bugs.launchpad.net/neutron/+bug/2020114 - [sqlalchemy-20] ``Engine`` class no longer inherits from ``Connectable`` Fix proposed: https://review.opendev.org/c/openstack/neutron/+/883521 Assigned to Rodolfo Wishlist ----------- - https://bugs.launchpad.net/neutron/+bug/2019960 - [RFE] Can't protect the "default" security group from regular users Unassigned Undecided -------------- - https://bugs.launchpad.net/neutron/+bug/2020001 - Neutron Dynamic Routing : vip is not advertised via BGP Marked incomplete by Jens Harbott so I decided to wait on their opinion - https://bugs.launchpad.net/neutron/+bug/2020060 - Stateless Feature of Security Group Not Functioning in Case of other Port same compute use stateful Rodolfo is discussing with the reporter more information about this but, since it is linux bridge driver I was not sure about confirming Assigned to Slawek - https://bugs.launchpad.net/neutron/+bug/2020168 - [OVN][SRIOV] traffic problems when sriov and non-sriov ports are bound on the same hypervisor The reporter is not sure if this is a bug in Neutron or in OVN. Someone with SRIOV knowledge might be able to help better. Unassigned Kind regards, Elvira

1 0

[kolla-ansible] Plugin pulling from other repositories
by James Leong 21 May '23

21 May '23

Hi everyone, How can I change the location where the repositories are being git pulled? I am using kolla-ansible as my deployment tool for deploying the yoga version of OpenStack. It seems like all plugins, such as zun_ui, blazar_dashboard, etc., are all being pulled from this path " https://opendev.org/openstack." I have attempted to change the variable kolla_dev_repos_git in globals.yml from "https://opendev.org/openstack" to point to my GitHub repository. However, it did not work. Is there a place where I can override the default path to pull plugins from my repository? Thanks for your help, James

2 1

Re: OVS-DPDK poor performance with Intel 82599
by Soheil D.Amiri 20 May '23

20 May '23

Dear satish About your topic on openstack.org by subject of "OVS-DPDK poor performance with Intel 82599". I have the same problem. I could not get good performance on my compute node. Did you solve your problem ? Would you please guide me about this problem https://lists.openstack.org/pipermail/openstack-discuss/2020-November/01912… Best Regards Soheil Amiri

1 0

[OSA][NEUTRON] Doubts about network setup
by Murilo Morais 19 May '23

19 May '23

Good evening everyone! I'm trying to set up a lab for testing using Openstack Ansible (OSA) and I'm having a lot of trouble understanding/setting up the network. I'm trying something similar to AIO (All-in-one) but with customizations (2 compute node). I'm using Debian 11 as OS. My problem is that I need the instances to communicate through VLANs that are being delivered directly to the interface of each compute node, as I need the same instances to participate in an existing network. I have a lot of doubts about this type of setup and how the configuration of provider_networks would be. Thanks in advance!

2 1

[kolla-ansible] Adding a USB device to a Docker container
by Bram Kranendonk 19 May '23

19 May '23

Hi OpenStack Discuss, I'm looking for a way to mount an USB device to a kolla-ansible Docker container. Is there a way to achieve this using the kolla-ansible configuration global vars? I couldn't find a way to do so. Thanks in advance, . Bram Kranendonk System Engineer <tel:> Oostmaaslaan 71 (15e etage) 3063 AN Rotterdam The Netherlands [team.blue logo]<https://team.blue/?utm_source=signature&utm_medium=email&utm_campaign=team-…>

1 0

[neutron] policy rules: filter on name field
by Paolo Emilio Mazzon 19 May '23

19 May '23

Hello, I'm trying to understand if this is feasible: I would like to avoid a regular user from tampering the "default" security group of a project. Specifically I would like to prevent him from deleting sg rules *from the default sg only* I can wite a policy.yaml like this # Delete a security group rule # DELETE /security-group-rules/{id} # Intended scope(s): project "delete_security_group_rule": "role:project_manager and project_id:%(project_id)s" but this is sub-optimal since the regular member can still *add* rules... Is it possible to create a rule like "sg_is_default" : ...the sg group whose name is 'default' so I can write "delete_security_group_rule": "not rule:sg_is_default" ? Thanks! Paolo -- Paolo Emilio Mazzon System and Network Administrator paoloemilio.mazzon[at]unipd.it PNC - Padova Neuroscience Center https://www.pnc.unipd.it Via Orus 2/B - 35131 Padova, Italy +39 049 821 2624

4 7

[cinder] festival of feature reviews 19 May 2023
by Rajat Dhasmana 19 May '23

19 May '23

Hello Argonauts, We will be having our monthly festival of reviews today i.e. 19th May (Friday) from 1400-1600 UTC. Following are some additional details: Date: 19th May, 2023 Time: 1400-1600 UTC Meeting link: will be shared in #openstack-cinder around 1400 UTC etherpad: https://etherpad.opendev.org/p/cinder-festival-of-reviews Thanks Rajat Dhasmana

1 0

[neutron] Drivers meeting
by Rodolfo Alonso Hernandez 19 May '23

19 May '23

Hello Neutrinos: Remember this afternoon we have the drivers meeting at 14UTC. The agenda has one topic: https://wiki.openstack.org/wiki/Meetings/NeutronDrivers As commented during the last team meeting, I also want you to review the open specs. Please spend some minutes on them: * https://review.opendev.org/q/project:openstack%252Fneutron-specs+status:open * https://review.opendev.org/c/openstack/nova-specs/+/859290 Regards.

1 0

[oslo.metrics][nova][oslo][oslo.messaging] oslo.metrics patch not working
by hai wu 18 May '23

18 May '23

I tried to backport this particular patch in order to add support for 'oslo.metrics' on openstack controller node first: https://github.com/openstack/oslo.messaging/commit/bdbb6d62ee20bfd5ffc59f87…. But it does not work. After adding the following configs in /etc/nova/nova.conf: [DEFAULT] default_log_levels = oslo.messaging=DEBUG [oslo_messaging_metrics] metrics_enabled = True I could only see messages related to 'oslo.messaging._drivers' in nova log files, and there's nothing related to 'oslo.messaging._metrics', or any other ones under 'oslo.messaging'. What else is needed for this back-ported patch to work on the controller?

1 0

[ironic] Core team updates
by Jay Faulkner 18 May '23

18 May '23

Hey all, I wanted to let you know that Shivanand Tendulker let me know they will no longer be working on Ironic. I've removed them from the ironic-core group as a result. As usual, with former Ironic core reviewers, Shivanand will be welcomed back quickly if they return to work on Ironic. Thank you for years of helping us keep things maintained! -- Jay Faulkner Ironic PTL TC Vice-Chair

1 0

[kolla] [train] haproxy and controller restart causes user impact
by Albert Braden 18 May '23

18 May '23

We have our haproxy and controller nodes on KVM hosts. When those KVM hosts are restarted, customers who are building or deleting VMs see impact. VMs may go into error status, fail to get DNS records, fail to delete, etc. The obvious reason is because traffic that is being routed to the haproxy on the restarting KVM is lost. If we manually fail over haproxy before restarting the KVM, will that be sufficient to stop traffic being lost, or do we also need to do something with the controller?

3 13

Re: 答复: [ptl] Need PTL volunteer for OpenStack Sahara
by Ghanshyam Mann 17 May '23

17 May '23

Thanks Jerry, Brin for the updates and showing interest in Sahara project. In PTG discussion, there were some concern on Sahara maintenance even we had PTL volunteer to lead this project but not much things got merged and gate was broken many times. But let me discuss it in TC meeting tomorrow and update here. -gmann ---- On Mon, 10 Apr 2023 00:22:08 -0700 Brin Zhang(张百林) wrote --- > Hi Gmann, > Jerry Zhou works the same company as qiujunting (Qiujunting has out of the office, so he cannot response us, I am so sorry), and he will be involved in Sahara. > Hope TC can add him in Sahara team, let him gradually complete the maintenance of the Sahara project. > Thanks. > > brinzhang > > -----邮件原件----- > 发件人: Jerry Zhou (周鑫勇) zhouxinyong(a)inspur.com> > 发送时间: 2023年4月7日 11:10 > 收件人: gmann(a)ghanshyammann.com; Juntingqiu Qiujunting (邱军婷) qiujunting(a)inspur.com> > 抄送: openstack-discuss(a)lists.openstack.org; Tom Guo(郭涛2810) guotao.bj(a)inspur.com>; Brin Zhang(张百林) zhangbailin(a)inspur.com>; Weiting Kong (孔维亭) kongweiting(a)inspur.com> > 主题: 答复: [ptl] Need PTL volunteer for OpenStack Sahara > > Hi Gmann, > I can lead this project. > > -----邮件原件----- > 发件人: Ghanshyam Mann gmann(a)ghanshyammann.com> > 发送时间: 2023年3月23日 2:35 > 收件人: Qiujunting qiujunting(a)inspur.com>; openstack-discuss openstack-discuss(a)lists.openstack.org> > 主题: [ptl] Need PTL volunteer for OpenStack Sahara > > Hi Qiu, > > I am reaching out to you as you were PTL for OpenStack Sahara project in the last cycle. > > There is no PTL candidate for the next cycle (2023.2), and it is on the leaderless project list. Please check if you or anyone you know would like to lead this project. > > - https://etherpad.opendev.org/p/2023.2-leaderless > > Also, if anyone else would like to help leading this project, this is time to let TC knows. > > -gmann > >

4 7

[Yoga][Magnum] change boot disk storage for COE VMs
by wodel youchi 17 May '23

17 May '23

Hi, When creating Magnum clusters, two disks are created for each VM sda and sdb, the sda is the boot disk and the sdb is for docker images. By default the sda is created (at least in my deployment) in nova vms pool, as ephemeral disks, the second disk sdb is created in cinder volume. Is it possible to move the sda from nova vms to cinder volume? I tried with default_boot_volume_type, but it didn't work. default_boot_volume_type = MyCinderPool Regards.

2 1

Installation openstack Multi node
by BEDDA Fadhel 17 May '23

17 May '23

Good morning, I am looking for a complete video or digital procedure that allows me to set up an openstack multi node test environment on vamware workstation. THANKS

4 3

[largescale-sig] Next meeting: May 17, 15utc
by Thierry Carrez 17 May '23

17 May '23

Hi everyone, The Large Scale SIG will be meeting this Wednesday in #openstack-operators on OFTC IRC, at 15UTC, our EU+US-friendly time. Kristin will be chairing. You can doublecheck how that UTC time translates locally at: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230517T15 Feel free to add topics to the agenda: https://etherpad.opendev.org/p/large-scale-sig-meeting Regards, -- Thierry Carrez

1 1

Transition to kolla
by Ryszard Mielcarek 17 May '23

17 May '23

Hello, is there anyone who already did transition on existing cluster, from package-version to kolla-ansible managed installation? Looking for guides, tips, your experience on that. BR, crazik

4 4

[ironic] RFC: In-Person-only PTG in Vancouver
by Jay Faulkner 17 May '23

17 May '23

Hey all, It was brought to my attention that there will not be much opportunity for remote participation in the PTG. We would have to manage the A/V ourselves in a shared room which would likely be extremely disruptive and ineffective. So I'm left with a few questions: should we continue forward with a PTG session for Ironic in Vancouver? What topics would be suitable with only a limited number of core reviewers available? I propose that the Ironic team uses the PTG venue for an informal collaboration and hacking space. It doesn't seem right for us to make any in-depth plans with only a subset of the core team. What do you all think? -- Jay Faulkner Ironic PTL TC Vice-Chair

4 3

Cinder Bug Report 2023-05-17
by Sofia Enriquez 17 May '23

17 May '23

Hello Argonauts, Cinder Bug Meeting Etherpad <https://etherpad.opendev.org/p/cinder-bug-squad-meeting> *Medium* - [RBD] Retyping of in-use boot volumes renders instances unusable (possible data corruption) <https://bugs.launchpad.net/cinder/+bug/2019190> - *Status:* Unassigned. *Low* - Volume upload to glance as image,use compression to accelerate gzip. Occasionally, there may be errors. <https://bugs.launchpad.net/cinder/+bug/2019549> - *Status:* Fix proposed to master <https://review.opendev.org/c/openstack/cinder/+/883147>. - [DELL Unity] Image volume creation fails in Unity. <https://bugs.launchpad.net/cinder/+bug/2019943> - *Status: *Unassigned. Cheers -- Sofía Enriquez she/her Software Engineer Red Hat PnT <https://www.redhat.com> IRC: @enriquetaso @RedHat <https://twitter.com/redhat> Red Hat <https://www.linkedin.com/company/red-hat> Red Hat <https://www.facebook.com/RedHatInc> <https://www.redhat.com>

1 0

[neutron] floating ip port forwarding and dvr
by 韩光宇 17 May '23

17 May '23

Hello, I know that "For instances with a floating IPv4 address, routing between self-service and provider networks resides completely on the compute nodes"[1] when enable neutron dvr. This can eliminate performance issues with network nodes. But it need fix floating ip to vm. I want to ask when employing floating IP port forwarding, is it feasible to enable direct access to the compute nodes for traffic? I may encounter a scenario where tens of thousands of ports are forwarded from a floating IP to different ports on different VMs. This would result in a significant concentration of traffic at the network node. I would like to inquire if there are any methods available to alleviate the pressure on the network node or to directly distribute the traffic across the compute nodes. Appreciate for any assistance or suggestions. Best regards, Han Guangyu [1] https://docs.openstack.org/neutron/train/admin/deploy-ovs-ha-dvr.html#deplo…

1 0

Train EOL
by Scott Little 17 May '23

17 May '23

Hello OpenStack community I'm one of the members of the StarlingX community. We've had a lot of stability issues with our ability to compile both our development branch and supported release branches these last few weeks. It all traces back to the Train EOL. We weren't monitoring openstack mailing lists, and missed the EOL announcement. We are actively moving off of Train, but aren't yet ready. What's really causing us grief is that some sub-projects, e.g.heat and nova, have started deleting elements of Train, e.g. git branches. Now please don't take this wrong. Ending support for an old branch is a totally normal thing, and we accept that. If StarlingX customers need support in that area, we'll provide it. However I would plea to you to NOT delete the elements of Train that allow other projects to compile old openstack releases, e.g. your gits branches. Sincerely Scott Little on behalf of StarlingX

5 7

Devstack and SQLAlchemy (Problems with stack.sh)
by Matt Watkins 17 May '23

17 May '23

Folks, I’m having problems running the devstack install script (stack.sh) under an Ubuntu-22.04.2 VM. The same point in the script is causing the install to abort, but the exact error depends on some Python dependencies. Anybody have a possible solution to this; it looks like some kind of Python dependency issue? Error with SQLAlchemy==1.4.48 +lib/keystone:init_keystone:489 /usr/local/bin/keystone-manage --config-file /etc/keystone/keystone.conf db_sync CRITICAL keystone [-] Unhandled error: sqlalchemy.exc.NoSuchModuleError: Can't load plugin: sqlalchemy.plugins:dbcounter ERROR keystone File "/home/devstack/.local/lib/python3.10/site-packages/sqlalchemy/util/langhelpers.py", line 343, in load ERROR keystone raise exc.NoSuchModuleError( ERROR keystone sqlalchemy.exc.NoSuchModuleError: Can't load plugin: sqlalchemy.plugins:dbcounter Error with SQLAlchemy==2.0.12 +lib/keystone:init_keystone:489 /usr/local/bin/keystone-manage --config-file /etc/keystone/keystone.conf db_sync Traceback (most recent call last): File "/usr/local/lib/python3.10/dist-packages/migrate/changeset/databases/sqlite.py", line 13, in <module> from sqlalchemy.databases import sqlite as sa_base ModuleNotFoundError: No module named 'sqlalchemy.databases' Thanks in advance, - Matt

4 5

[TripleO] Removal of TripleO Zed Integration and Component Lines
by Ronelle Landy 16 May '23

16 May '23

Hello All, Removal of TripleO Zed Integration and Component Lines Per the decision to not maintain TripleO after the Zed release [1], the Zed integration and component lines are being removed in the following patches: https://review.rdoproject.org/r/c/config/+/48073 https://review.rdoproject.org/r/c/config/+/48074 https://review.rdoproject.org/r/c/rdo-jobs/+/48075 To be clear, please note that following these changes, the gate for stable/zed TripleO repos will no longer be updated or maintained. Per the earlier communications, there should be no more patches submitted for stable/zed TripleO repos, and any backports will go to stable/wallaby or stable/train. The last promoted release of Zed through TripleO is: https://trunk.rdoproject.org/centos9-zed/current-tripleo/delorean.repo (hash:61828177e94d5f179ee0885cf3eee102), which was promoted on 05/15/2023. The Ceph promotion lines related to Zed are also removed in the above patches. Check/gate testing for the master branch is in process of being removed as well (https://review.opendev.org/c/openstack/tripleo-ci/+/882759) [1] https://review.opendev.org/c/openstack/governance/+/878799

5 8

[Glance] Weekly Meeting Cancelled
by Pranali Deore 16 May '23

16 May '23

Hello, As per the discussion [1], glance weekly meeting is cancelled for this week as most of the team members will not be around. For next week - 25th May, we will have the meeting if anything comes urgent, otherwise we'll directly meet on Thursday 1st June. [1]: https://meetings.opendev.org/meetings/glance/2023/glance.2023-05-11-13.59.l… Thanks & Regards, Pranali Deore

1 0

[keystone][horizon][kolla-ansible] user access specific domain
by James Leong 16 May '23

16 May '23

Hi all, I am playing around with the domain in the yoga version of OpenStack using kolla-ansible as the deployment tool. I have set up Globus as my authentication tool. However, I am curious if it is possible to log in to an existing OpenStack user account via federated login (based on Gmail) In my case, first, I created a user named "James" in one of the domains called federated_login. When I attempt to log in, a new user is created in the default domain instead of the federated_login domain. Below is a sample of my globus.json. [{"local": [ { "user": { "name":"{0}, "email":"{2} }, "group":{ "name": "federated_user", "domain: {"name":"{1} } } ], "remote": [ { "type":"OIDC-name"}, { "type":"OIDC-organization"},{"type":"OIDC-email"} ] }] Apart from the above question, is there another easier way of restricting users from login in via federated? For example, allow only existing users on OpenStack with a specific email to access the OpenStack dashboard via federated login. Best Regards, James

2 3

[Storyboard] How to reach the Storyboard operators / maintainers?
by Christian Rohmann 16 May '23

16 May '23

Hello, a while back I opened a story / bug on storyboard (https://storyboard.openstack.org/#!/story/2010689) about improving email deliver-ability. I am uncertain if that was the right place to raise this issue or if there is another channel. In any case, could someone kindly point me or this suggestion into the right direction? Regards Christian

2 1