- openstack-discuss - lists.openstack.org

[zuul][neutron][neutron-fwaas][tap-as-a-service] Errors in the jobs definitions and EOL of some old neutron-fwaas and tap-as-a-service branches
by Lajos Katona 11 Apr '22

11 Apr '22

Hi, During the PTG we got questions about projects which cause zuul config errors (see [1]). Slawek started a thread regarding networking-midonet (see [2]), I join to that with neutron-fwaas and tap-as-a-servie. - neutron-fwaas' stable/rocky and stable/queens branches - tap-as-a-service stable/ocata, stable/pike, stable/queens and stable/rocky branches are affected - as tap-as-a-service was under x/ namespace previously I am not sure what can happen to those branches. If there is nobody to fix the config and gate errors on these old branches, then the best would be to EOL and later delete these branches. [1]: https://zuul.opendev.org/t/openstack/config-errors [2]: http://lists.openstack.org/pipermail/openstack-discuss/2022-April/028120.ht…

1 0

[requirements][kolla][infra] Python versions in u-c
by Marcin Juszkiewicz 11 Apr '22

11 Apr '22

OpenStack Zed lists Python 3.8 as minimal version. Debian 'bullseye' (current stable) as Python 3.9, Ubuntu 20.04 'focal' has 3.8. Ubuntu 22.04 'jammy' has 3.10, CentOS Stream 9 (so RHEL 9, Rocky Linux 9) has 3.9 version. We ignore anything RHEL 8 based as they have Python 3.6 only and there is no work on getting OpenStack Zed supported there. So there will be 3.8, 3.9, 3.10 used with Zed. Two question to requirements team: 1. Are there plans to get rid of 3.6 entries from openstack/requirements/upper-constraints.txt in near future? 2. Can we get "python_version=='3.8'" lines be changed to ">=" ones? Or once 3.6 entries drop just assume that we have 3.8+ and drop 'python_version' entries? websocket-client===1.3.1;python_version=='3.6' websocket-client===1.3.2;python_version=='3.8' Asking due to my recent Kolla/infra work on getting CentOS Stream 9 supported in both places. Infra team has CI job which builds Python wheel cache to make sure that other CI jobs do not have to do it. Times are nicely cut due to this stuff. There are 55 entries with "python_version=='3.8'". Probably none of them are time critical anymore as upstream projects provide aarch64 wheels most of times too.

2 1

[Zuul] [neutron] Errors in the jobs definitions and EOL of some old networking-midonet branches
by Slawek Kaplonski 11 Apr '22

11 Apr '22

Hi, Today I got back to checking my old patches which are going to fix some zuul jobs' definitions. And I found that CI for networking-midonet in some stable branches is totally broken: * stable/ussuri: https://review.opendev.org/c/openstack/networking-midonet/+/823273[1] - everything is red here * stable/train: https://review.opendev.org/c/openstack/networking-midonet/+/823275[2] - most of the jobs are red * stable/stein: https://review.opendev.org/c/openstack/networking-midonet/+/823276[3] - here there were errors in the definition of the jobs, I changed it now, maybe it will run, So my question is if we want, and have resources, to fix it somehow or should we maybe set those branches as EOL now? Also, another question to the Zuul team - will it be ok if such branch will be EOL to get rid of the Zuul configuration errors warnings? Or should we somehow force to merge those patches first to get rid of the errors in https://zuul.opendev.org/t/openstack/config-errors[4] ? -- Slawek Kaplonski Principal Software Engineer Red Hat -------- [1] https://review.opendev.org/c/openstack/networking-midonet/+/823273 [2] https://review.opendev.org/c/openstack/networking-midonet/+/823275 [3] https://review.opendev.org/c/openstack/networking-midonet/+/823276 [4] https://zuul.opendev.org/t/openstack/config-errors

2 2

[all][neutron] Pike - End of Life
by Lajos Katona 11 Apr '22

11 Apr '22

Hi, As time and cycles are passing it's time to bring up the topic of EOLing an old branch again. In Neutron (and its satellite projects, aka stadium) stable/pike has had low activity in the last cycle and recently even the jobs for it started to fail, and we cause some zuul config errors which are hard to fix as the gate is broken. Based on the Ocata EOL mail (see [1]), Neutron team decided to EOL Pike, see the discussions during the Zed PTG (see [2]) [1]: http://lists.openstack.org/pipermail/openstack-discuss/2021-April/021949.ht… [2]: https://etherpad.opendev.org/p/neutron-zed-ptg#L254 Lajos Katona (lajoskatona)

1 0

[vmware-nsx] EOL old branches
by Slawek Kaplonski 11 Apr '22

11 Apr '22

Hi Salvatore, According to Your comments in my patches to x/vmware-nsx project: * stable/stein: https://review.opendev.org/c/x/vmware-nsx/+/823277[1] * stable/rocky: https://review.opendev.org/c/x/vmware-nsx/+/823291[2] Can You maybe make them EOL finally? -- Slawek Kaplonski Principal Software Engineer Red Hat -------- [1] https://review.opendev.org/c/x/vmware-nsx/+/823277 [2] https://review.opendev.org/c/x/vmware-nsx/+/823291

1 0

[neutron] Bug Deputy Report April 04 - 11
by Lucas Alvares Gomes 11 Apr '22

11 Apr '22

Hi, This is the Neutron bug report from April 4th to 11th. Critical: * https://bugs.launchpad.net/neutron/+bug/1967893 - " [stable/yoga] tempest.scenario.test_network_qos_placement.MinBwAllocationPlacementTest fails in neutron-ovs-tempest-multinode-full job Edit" - Assigned to: Lajos Katona High: * https://bugs.launchpad.net/neutron/+bug/1967839 - " [L3] NDP extension not handing "ha_state_change" correctly" - Assigned to: Rodolfo Alonso * https://bugs.launchpad.net/neutron/+bug/1967996 - "[OVN] External subscribers to feed generated in Openstack fails if no internal VM is subscribed to that feed first" - Unassigned Medium: * https://bugs.launchpad.net/neutron/+bug/1967742 - "ML2 - Network Context, not possible to see original/current segments" - Unassigned * https://bugs.launchpad.net/neutron/+bug/1968343 - "Security Group Rule create with forged integer security_group_id causes exceptions" - Assigned to: Andrew Karpow Needs further triage: * https://bugs.launchpad.net/neutron/+bug/1968057 - "Doc needed for configuring neutron AZs for edge" - Unassigned * https://bugs.launchpad.net/neutron/+bug/1968269 - " [router qos] qos binding is not clear after remove gateway" - Unassigned Cheers, Lucas

1 0

[Senlin][Zuul] Errors in the jobs definitions in stable branches
by Slawek Kaplonski 11 Apr '22

11 Apr '22

Hi, Some time ago I pushed patches to fix definitions of the Zaqar's jobs in stable branches: * stable/stein: https://review.opendev.org/c/openstack/senlin/+/823287[1] * stable/rocky: https://review.opendev.org/c/openstack/senlin/+/823293[2] * stable/queens: https://review.opendev.org/c/openstack/senlin/+/823294[3] It seems that CI for all those branches is broken. Do You still have plans to fix it so I can rebase my patches and we can get rid of those zuul's configuration errors? Or should those branches be maybe EOL? -- Slawek Kaplonski Principal Software Engineer Red Hat -------- [1] https://review.opendev.org/c/openstack/senlin/+/823287 [2] https://review.opendev.org/c/openstack/senlin/+/823293 [3] https://review.opendev.org/c/openstack/senlin/+/823294

1 0

[Zaqar][Zuul][TC] Errors in the jobs definitions in stable branches
by Slawek Kaplonski 11 Apr '22

11 Apr '22

Hi, Some time ago I pushed patches to fix definitions of the Zaqar's jobs in stable branches: * stable/stein: https://review.opendev.org/c/openstack/zaqar/+/823289[1] * stable/rocky: https://review.opendev.org/c/openstack/zaqar/+/823295[2] * stable/queens: https://review.opendev.org/c/openstack/zaqar/+/823296[3] * stable/pike: https://review.opendev.org/c/openstack/zaqar/+/823297[4] It seems that CI for all those branches is broken. Do You still have plans to fix it so I can rebase my patches and we can get rid of those zuul's configuration errors? Or should those branches be maybe EOL? I'm tagging TC in the topic of that email too as there is still no PTL of the Zaqar project appointed. -- Slawek Kaplonski Principal Software Engineer Red Hat -------- [1] https://review.opendev.org/c/openstack/zaqar/+/823289 [2] https://review.opendev.org/c/openstack/zaqar/+/823295 [3] https://review.opendev.org/c/openstack/zaqar/+/823296 [4] https://review.opendev.org/c/openstack/zaqar/+/823297

1 0

error
by Reza Bojnordi 11 Apr '22

11 Apr '22

Hi, sorry I have problem on Openstack tcd3gw': No module named 'etcd3gw': ModuleNotFoundError: No module named 'etcd3gw' Apr 09 20:09:34 infra1-nova-api-container-483c43d7 nova-metadata-wsgi[2912]: 2022-04-09 20:09:34.836 2912 ERROR stevedore.extension [req-089ca62f-02ee-49f9-a305-75f426102ebc - - - - -] Could not load 'oslo_cache.e tcd3gw': No module named 'etcd3gw': ModuleNotFoundError: No module named 'etcd3gw' -- <https://about.me/rbojnordi?promo=email_sig&utm_source=product&utm_medium=em…> reza bojnordi about.me/rbojnordi <https://about.me/rbojnordi?promo=email_sig&utm_source=product&utm_medium=em…>

4 4

[baremetal-sig][ironic] Tue Apr 12, 2022, 2pm UTC: "Using OpenStack Ironic for HPC at Berlin Institute of Health"
by Arne Wiebalck 11 Apr '22

11 Apr '22

Dear all, The Bare Metal SIG will meet tomorrow Tue Apr 12, 2022, 2pm UTC featuring a topic of the day presentation by Dr. Manuel Holtgrewe: "Bare Metal for Health - Using OpenStack Ironic for HPC at Berlin Institute of Health" Everyone is welcome, all details on how to join can be found on the SIG's etherpad: https://etherpad.opendev.org/p/bare-metal-sig Hope to see you there! Arne -- Arne Wiebalck CERN IT

1 0