April 2022 - openstack-discuss - lists.openstack.org

[swift][ptg] Ops feedback session - Apr 7 at 13:00 UTC
by Timothy Burke 06 Apr '22

06 Apr '22

As in PTGs past, we're getting devs and ops together to talk about Swift: what's working, what isn't, and what would be most helpful to improve. We're meeting in Havana (https://www.openstack.org/ptg/rooms/havana) on Apr 7 at 13:00UTC -- if you run a Swift cluster, we hope to see you there! Even if you can't make it, I'd appreciate if you can offer some feedback on this PTG's etherpad (https://etherpad.opendev.org/p/swift-zed-ops-feedback) Tim

3 2

[ptg][ptl][largescale-sig][all] PTG session: "The Scaling Journey" - Wednesday, April 6 15utc - kilo room
by Belmiro Moreira 06 Apr '22

06 Apr '22

Hi, the Large Scale SIG is organizing a PTG session to discuss "The Scaling Journey". The SIG worked in a "scaling journey" to guide and help operators to scale their OpenStack deployments. Definitely, there are different ways to scale OpenStack! and the challenges to move from a few hundreds cores to thousands or now millions cores are completely different. Based on the experience of several operators, we tried to answer different common questions and identify the pain points. https://wiki.openstack.org/wiki/Large_Scale_SIG If you are interested in scalability it would be great to have your feedback. Also, it's important that PTLs join because they can give the project vision for scalability and advise on how to overcome possible bottlenecks. To discuss all of this we will have a "zoom" session on Wednesday, April 6 15utc - "kilo room". See you there! cheers, Belmiro on behalf of the Large Scale SIG

2 1

Upgrading Openstack nodes
by Marc-Antoine Godde 06 Apr '22

06 Apr '22

Hello, We are running an Openstack cloud composed of 3 controller nodes and 4 compute nodes. Our deployment was realized with OpenStack-ansible and we are running OpenStack Ussuri on Ubuntu 18.04. Our plan is to upgrade nodes to Ubuntu 20.04, that way we would be able to update to OpenStack Victoria and further. We would like to withdraw each node from the cluster, reinstall a clean linux and redeploy the nodes. There is garbage remaining from previous upgrades. We figured out the way in the documentation to remove a compute node from the cluster with Openstack-ansible but we can’t find any related documentation for controller nodes. Any help would be very much appreciated. By the way, if you’d have any other suggestions on how to perform that upgrade, fell free to help. Best, Marc-Antoine Godde

2 2

[dev][infra][qa][tact-sig] Zuul behavior change with Depends-On across queues
by Jeremy Stanley 06 Apr '22

06 Apr '22

For those who haven't seen the more detailed announcement[*] about it, just a quick note that if you get a sudden -2 back from Zuul when approving a change with a Depends-On to a change in a different project which hasn't merged yet, that's likely an indication those projects don't share a dependent queue. It's not a bug, but an intentional clarification of Zuul's enqueuing behavior. For most other Zuul deployments (and even our other Zuul tenants in OpenDev) this is purely cosmetic, but since OpenStack's Zuul tenant is configured to require a positive Verified vote before enqueuing into the gate pipeline, it means some changes may end up unexpectedly needing another pass through check first. It's worth re-evaluating whether or not this "clean check" rule remains a useful requirement for gating. It was added some years ago because a number of gate breaking bugs were traced back to unstable changes being rechecked enough times that eventually they got lucky and were able to merge, and then their instability contributed to destabilizing the integrated gate as a whole. Similarly, changes were being approved without reviewers confirming their jobs were passing first, and this led to additional resource waste. There is a bit of discussion around "blind rechecks" at the PTG this week, and so this topic is related; it might be a good idea to consider it in conjunction with the greater recheck conversation. [*] http://lists.opendev.org/pipermail/service-announce/2022-April/000033.html -- Jeremy Stanley

2 2

[openstack-ansible][PTG] Session on 6th of April is cancelled
by Dmitriy Rabotyagov 06 Apr '22

06 Apr '22

1 0

[Oslo] IRC meeting.
by Daniel Mats Niklas Bengtsson 06 Apr '22

06 Apr '22

Hi there, I would like to know if you agree that we have the meeting once a week? Instead of the first and third Monday of the month. It will be easier to manage and if sometimes it is canceled it does not matter Even if the meetings each week are short, this way we will have regular follow-up.

2 1

Metadata service {devices list}
by Ahmed Abdelhamid 06 Apr '22

06 Apr '22

Hi All, I am running into a strange issue with the metadata service. Per metadata-service manual <https://docs.openstack.org/nova/latest/user/metadata.html>, the devices attached to a VM should be visible in $ curl http://169.254.169.254/openstack/2018-08-27/meta_data.json However, whenever i execute it from a VM , the devices array is empty and looks like this, any idea why ? Thanks { "random_seed": "yu5ZnkqF2CqnDZVAfZgarG...", "availability_zone": "nova", "keys": [ { "data": "ssh-rsa AAAAB3NzaC1y...== Generated by Nova\n", "type": "ssh", "name": "mykey" } ], "hostname": "test.novalocal", "launch_index": 0, "meta": { "priority": "low", "role": "webserver" }, "devices": [ ], "project_id": "f7ac731cc11f40efbc03a9f9e1d1d21f", "public_keys": { "mykey": "ssh-rsa AAAAB3NzaC1y...== Generated by Nova\n" }, "name": "test"}

3 10

wireguard - ? - puzzle
by lejeczek 06 Apr '22

06 Apr '22

Hi guys. Has anybody solved that puzzle? Or perhaps it's not a puzzle at all, I'd imagine might be trivial to experts. First I thought - and only thought so far thus asking here - 'allowed_address_pairs' I'd need but that obviously does not do anything as 'wireguard' creates its own ifaces. So.. how do you get your 'wireguard' in openstack to route (no NAT) to instances' local network(s)? many thanks, L.

3 3

[Octavia] Weekly meeting cancelled
by Gregory Thiemonge 05 Apr '22

05 Apr '22

Hi, As this is the PTG week, the Octavia weekly meeting is cancelled. Thanks, Gregory

1 0

[tripleo][RDO] Tags for Yoga GA version in tripleo projects
by Alfredo Moralejo Alonso 05 Apr '22

05 Apr '22

Hi, As we are doing builds for RDO Yoga release, I'd like to know what we should include in yoga release. In Xena, tags were created with the commits for the most recent promotion at around GA and shipped rpm packages for them. Additionally, we pinned RDO Trunk Xena to those versions. Will we get similar tags for Yoga? Best regards, Alfredo

2 3