May 2020 - openstack-discuss - lists.openstack.org

[neutron] Recent incompatible change in stable/rocky branch
by Neil Jerram 19 May '20

19 May '20

I'm sorry, but this is a moan. This merge - https://opendev.org/openstack/neutron/commit/a6fb2faaa5d46656db9085ad6bcfc6… - to the Neutron stable/rocky branch on April 23rd, has broken my team's Neutron plugin, by requiring 3rd party LinuxInterfaceDriver subclasses to take a new 'link_up' argument in their 'plug_new' method. IMO, it should have been obvious to folk proposing or reviewing this, that it would cause breakage. Does Neutron have a different understanding of "stable" than I do? Or do plugins other than OVN not matter anymore?

2 5

[neutron] WSREP: referenced FK check fail: Lock wait index
by Fabian Zimmermann 18 May '20

18 May '20

Hi, im just upgrading my galera nodes from 5.6 to 5.7 and pike -> queens. Now im seeing this lines hitting my mysql.log: -- 2020-05-19T05:41:39.516879Z 11861 [ERROR] InnoDB: WSREP: referenced FK check fail: Lock wait index `PRIMARY` table `neutron`.`provisioningblocks` -- Is this a known issue? Anything I may/should do? Thank a lot, Fabian

1 1

[ironic] III SPUC - The Sanity Preservation Un-Conference
by Iury Gregory 18 May '20

18 May '20

Hello everyone \o/ It's time for the third edition of SPUC! It will happen this Friday (May 22) at 5pm UTC! For those who doesn't know what is SPUC check [0]. We will be using Julia's bluejeans: https://bluejeans.com/u/jkreger Etherpad for <crazy> ideas: https://etherpad.opendev.org/p/IIISanityPreservationUnConference [0] http://lists.openstack.org/pipermail/openstack-discuss/2020-March/013521.ht… -- *Att[]'sIury Gregory Melo Ferreira * *MSc in Computer Science at UFCG* *Part of the puppet-manager-core team in OpenStack* *Software Engineer at Red Hat Czech* *Social*: https://www.linkedin.com/in/iurygregory *E-mail: iurygregory(a)gmail.com <iurygregory(a)gmail.com>*

2 1

[all] Virtual Ussuri Celebration
by Kendall Nelson 18 May '20

18 May '20

Hello Everyone! I wanted to invite you all to a virtual celebration of the release! Next Friday, May 15th at 20:00 UTC, I invite you to join me in celebrating :) The purpose of this meeting will be two fold: 1. To celebrate all that our community has accomplished over the last release since we can't get together in person in June. I was thinking with trivia and a baking contest (I was going to attempt a cake in the shape of the OpenStack logo or maybe the actual Ussuri logo) :) There was also a request for piano karaoke which is also still on the table.. 2. To test out the OpenDev team's meetpad instance and work out kinks so that it can be vetted for PTG use. Here is the room we will be testing: https://meetpad.opendev.org/virtual-ussuri-celebration Worst case scenario, I'll share my zoom. Can't wait to see you all there (and what you've baked)! -Kendall (diablo_rojo) <https://meetpad.opendev.org/virtual-ussuri-celebration>

3 4

[tc][tricircle] Retiring the Tricircle project
by Ghanshyam Mann 18 May '20

18 May '20

Hello Everyone, As you know, Tricirlce is a leaderless project for the Victoria cycle means there is no PTL candidate for Victoria cycle. 'No PTL' is one of the criteria which triggers TC to start checking the health, maintainers of the project for dropping project from OpenStack Governance[1]. TC discussed the Tricircle project in today's meeting[2] and checked if the project has maintainers but no PTL or it is missing the maintainer completely. By looking at the Ussuri cycle development work, it seems no activity related to Tricircle except few gate fixes or community goal commits[2]. Based on all these checks, we are going to start the process of removing Tricircle from governance in the Victoria cycle as specified in the Mandatory Repository Retirement resolution [4] and detailed in the infra manual [5]. NOTE: release team is trying to do Ussuri release for Tricircle's deliverables if we can get this reno fix merge - https://review.opendev.org/#/c/721697/ >From the Victoria cycle, Tricircle will move out of OpenStack governance by keeping their repo under OpenStack namespace with an empty master branch with 'Not Maintained' message in README. If someone from old or new maintainers shows interest to continue its development then it can be re-added to OpenStack governance. With that thanks to all Tricircle contributors and PTLs for maintaining this project. [1] https://governance.openstack.org/tc/reference/dropping-projects.html [2] http://eavesdrop.openstack.org/meetings/tc/2020/tc.2020-04-21-15.01.log.html [3] https://www.stackalytics.com/?metric=commits&release=ussuri&module=tricircl… [4] https://governance.openstack.org/tc/resolutions/20190711-mandatory-reposito… [5] https://docs.opendev.org/opendev/infra-manual/latest/drivers.html#retiring-… -gmann

1 1

[nova][ptg] Ussuri Retrospective
by Balázs Gibizer 18 May '20

18 May '20

Hi, There is a common but seemingly not that popular topic on each nova PTG, the retrospective. I added the retrospective to the PTG etherpad [0]. Please collect retrospective topics there or tell me privately if you have a sensitive topic you want me to bring up. If there won't be any topic collected until the PTG then we will simply skip the retrospective. Cheers, gibi [0] https://etherpad.opendev.org/p/nova-victoria-ptg

1 0

[blazar] IRC meeting cancelled on May 19
by Pierre Riteau 18 May '20

18 May '20

Hello, Due to a scheduling conflict, I will not be able to chair the Blazar IRC meeting on Tuesday May 19. I propose to cancel and meet as usual on May 26, when we will finalise the agenda for the PTG meeting. Best wishes, Pierre Riteau (priteau)

1 0

[neutron] bug deputy report for week of 2020-05-11
by Bence Romsics 18 May '20

18 May '20

Hi Neutrinos, Here's the deputy report for last week: High: * https://bugs.launchpad.net/neutron/+bug/1878042 SRIOV agent does not parse correctly "ip link show <dev>" fix proposed: https://review.opendev.org/726918 * https://bugs.launchpad.net/neutron/+bug/1878160 [OVN] Functional tests environment is using old OVN fix proposed (wip): https://review.opendev.org/727193 Medium: * https://bugs.launchpad.net/neutron/+bug/1877977 [DVR] Recovery from openvswitch restart fails when veth are used for bridges interconnection unassigned * https://bugs.launchpad.net/neutron/+bug/1875981 Admin deleting servers or ports leaves orphaned DNS records fix proposed (wip): https://review.opendev.org/728385 related test: https://review.opendev.org/728409 * https://bugs.launchpad.net/neutron/+bug/1878299 [floatingip port_forwarding] changing external port to used value hangs with retriable exception unassigned * https://bugs.launchpad.net/neutron/+bug/1878358 OVN migration doesn't clean up the existing ml2-ovs agents fix proposed: https://review.opendev.org/727648 * https://bugs.launchpad.net/neutron/+bug/1878632 Race condition in subnet and segment delete: The segment is still bound with port(s) assigned, no fix proposed yet test reproducing the bug: https://review.opendev.org/728904 * https://bugs.launchpad.net/neutron/+bug/1878916 When deleting a network, delete the segment RP only when the segment is deleted fix proposed: https://review.opendev.org/728507 Incomplete: * https://bugs.launchpad.net/neutron/+bug/1877978 SNAT Problem Floating IP * https://bugs.launchpad.net/neutron/+bug/1878719 DHCP Agent's iptables CHECKSUM rule causes skb_warn_bad_offload kernel * https://bugs.launchpad.net/neutron/+bug/1879009 attaching extra port to server raise duplicate dns-name error See you on the meeting, Bence

1 0

[nova][gate] openstack-tox-pep8 job failing
by melanie witt 18 May '20

18 May '20

Hey all, The openstack-tox-pep8 job is now failing too since hacking 3.1.0 was released [1], which brings in a newer flake8 version, which adds new checks to the code and we have a few failures. We have a change that would pin hacking >= 3.0.1 and < 3.1.0 [2] to prevent new checks from rolling in automatically, but it is stuck behind the failing openstack-tox-docs job [3]. This is unfortunately a chicken and egg situation. The proposed nova patch for the docs job, as mentioned from my other ML post: https://review.opendev.org/727898 fails the pep8 job and the proposed nova patch for the pep8 job [2] fails the docs job. I'm not sure how people will want to proceed here because a nova-side-only gate fix would involve squashing ^ and [2]. If we can get another fix or revert to openstackdocstheme to take care of the docs job, we'd then only need [2] on the nova side to unblock the nova gate. Cheers, -melanie [1] https://review.opendev.org/728016 [2] https://review.opendev.org/727347 [3] http://lists.openstack.org/pipermail/openstack-discuss/2020-May/014876.html

3 5

[gate] parallel releasenotes job builds
by melanie witt 18 May '20

18 May '20

Hi all, We recently merged a change [1] that will run the build-openstack-releasenotes job with 'sphinx-build -j auto' to allow the build to run on multiple CPUs. With the recent releases of reno 3.1.0 and openstackdocstheme 2.1.2, both libraries advertise that they are parallel_read_safe and thus can support the '-j' option. This effort was inspired by the nova build-openstack-releasenotes job which typically takes 52m to > 1h to run. The job timeout is 1h, so we were occasionally seeing TIMED_OUT job failures. With parallel sphinx-build, the nova job run time is reduced down to about 15m (awesome!). I wanted to say thanks to smcginnis, dhellmann, clarkb, and AJaeger for helping this newbie to reno and docs builds. If anyone encounters any issues around the parallel reno builds in their projects, please let us know. Cheers, -melanie [1] https://review.opendev.org/727473

3 2