February 2024 - openstack-discuss

[cinder][kubernetes] pvc stuck in reserved state
by Eugen Block 28 Feb '24

28 Feb '24

Hi, we have a couple of k8s clusters in our openstack environment (Victoria) and are facing some problems when pods get moved to different worker nodes. The PVCs get stuck in "reserved" state, and they only get attached after resetting the volumes to "available" state. First I thought it was due to the missing multiattach volume type which seems to work fine in our customer's cluster. I also found the "duplicate connector" error messages in the nova-compute.log: > Unable to update attachment.(Invalid volume: duplicate connectors > detected on volume But after adding the multiattach volume type and retyping the affected volumes, the PVCs still get stuck in "reserved" state while I don't see any error messages on the hypervisors anymore. So it seems that something has changed/improved, but now I don't have a trace where to look next. Could this be a cinder-csi driver issue or something? Our cinder backend is ceph (Pacific). Not sure which information could help so please ask for anything I can provide. Any pointers are highly appreciated! Thanks! Eugen

1 1

[ptl][election][kuryr] PTL non-candidacy for 2024.2
by Roman Dobosz 28 Feb '24

28 Feb '24

Hello, I'm not going to run for PTL of Kuryr for Dalmation. I encourage Kuryr contributors to run for PTL and will offer any assistance needed for whoever is elected. Thanks, Roman Dobosz (gryf)

1 0

[election][tc] Candidacy for the TC
by James Page 28 Feb '24

28 Feb '24

Hi All, I'd like to announce my candidacy for the Technical Committee. For those of you who do not know me, I've been involved with OpenStack from the very early days of the project. I've attended many OpenStack summits and Project Teams Gatherings over the years and have been privileged to help the project grow from its small beginnings to the established cloud platform that it has now become. I was the main driver behind migration of the OpenStack Charms to its current home as an official OpenStack project which I then led for many cycles as PTL and I remain an active core developer on this project. I'm the current PTL of Sunbeam which was accepted into OpenStack last year as a new incubating project, providing a reboot of the charmed approach to deploying OpenStack utilising the many new projects that have appeared since we started the OpenStack Charms project back in 2011. I also work as part of the team responsible for the packaging and distribution of OpenStack as part of Ubuntu. I have been a member of the TC for the last year providing a voice from a deployment tooling and distribution perspective. I would like the opportunity to continue in this role providing what I believe is a valuable perspective to the technical governance of the OpenStack project. In particular I think we need to address some of the challenges around the general health of wider ecosystem dependencies of OpenStack to ensure that we move forward onto a stable base benefiting operators, distributors and developers. Thank you for taking the time to consider my candidacy for this role. Regards James Page (irc:jamespage)

1 0

[puppet][murano] Retiring puppet-murano
by Takashi Kajinami 28 Feb '24

28 Feb '24

Hello, As you may know, TC decided to exclude Murano from 2024.1 release because of its health status[1]. Based on this fact, and also another fact that puppet-murano has had no real development activities for long, I'd propose retiring puppet-murano now. It may be too late to make release deliverable decision. However our puppet modules are tightly coupled with openstack release cycles and we can't assure compatibility between our modules and inactive components, which are no longer aligned with the consistent release cycle. Please let me know if you have any concerns. Thank you, Takashi [1] https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack… -- Takashi Kajinami irc: tkajinam github: https://github.com/kajinamit launchpad: https://launchpad.net/~kajinamit

1 0

[ptl][election][sunbeam] PTL candidacy for 2024.2
by Hemanth N 28 Feb '24

28 Feb '24

Hello OpenStack Community, I am nominating myself for the Sunbeam PTL position during the 2024.2 cycle. I am a core team member of the Sunbeam project from the incubation stage and contributed to most of the charms that are part of the sunbeam project. In the last cycle we focussed on moving the 20+ charm repo's to a mono repo which helped reduce the codebase and easy maintenance of all the sunbeam charms. In the 2024.2 cycle, my priorities will be * Adding new features or charms and filling gaps to enable charms to be production ready * Stabilise the existing charms or libraries * Bug smashing * CI stability Thank you for your support. Best Regards, Hemanth Nakkina (hemanth, hemanth.nakkina(a)canonical.com)

1 0

Tim Burke candidacy for Swift PTL
by Tim Burke 28 Feb '24

28 Feb '24

I'm pleased to announce my candidacy for Swift PTL for the Dalmation cycle. As usual, we have a great many projects currently in flight: - Supporting Python 3.12 - Adding more fine-grained metrics, while keeping it configurable to not break existing metrics pipelines nor overwhelm them - Improving S3 compatibility, enabling a broader community of tools to work out of the box - More operator tooling for expiring objects - Better process management for WSGI servers - Backend ratelimiting to improve per-node load shedding I'm excited to see how each of these will move forward over the next cycle. Tim Burke

2 1

[all][elections][ptl][tc] Combined PTL/TC 2024.2 cycle Election Nominations Last Days
by Ian Y. Choi 28 Feb '24

28 Feb '24

A quick reminder that we are in the last hours for declaring PTL and TC candidacies. Nominations are open until Feb 28, 2024 23:45 UTC. If you want to stand for election, don't delay, follow the instructions at[1] to make sure the community knows your intentions. Make sure your nomination has been submitted to the openstack/election repository and approved by election officials. All members of the community are encouraged to confirm their email address and opt in to the CIVS election as described at: https://governance.openstack.org/election/#electorate Election statistics[2]: Nominations started @ 2024-02-14 23:45:00 UTC Nominations end @ 2024-02-28 23:45:00 UTC Nominations duration : 14 days, 0:00:00 Nominations remaining : 23:49:45 Nominations progress : 92.91% --------------------------------------------------- Projects[1] : 47 Projects with candidates : 30 ( 63.83%) Projects with election : 0 ( 0.00%) --------------------------------------------------- Need election : 0 () Need appointment : 17 (Barbican Cloudkitty Ec2_Api Freezer Glance Keystone Kuryr Magnum Murano OpenStackAnsible OpenStack_Charms Sahara Senlin Skyline Solum Trove Watcher) --------------------------------------------------- TC Candidates : 7 (5 Required) =================================================== Stats gathered @ 2024-02-27 23:55:15 UTC This means that with approximately 2 days left, 17 projects will be deemed leaderless. In this case the TC will oversee PTL selection as described by [3]. Thank you, [1] https://governance.openstack.org/election/#how-to-submit-a-candidacy [2] Any open reviews at https://review.opendev.org/q/project:openstack/election+is:open+file:candid… have not been factored into these stats. [3] https://governance.openstack.org/resolutions/20141128-elections-process-for…

1 0

[ptl][election][cinder] PTL candidacy for 2024.2
by Jon Bernard 28 Feb '24

28 Feb '24

Hi All, I would like to announce my candidacy for PTL of the Cinder project for the 2024.2 cycle. I think Rajat has done an excellent job serving the project and I would like to carry existing efforts forward while continuing to focus maintaining a healthy and productive community around the Cinder project. I've been contributing to the Cinder project in varying capacity for close to 10 years. I've worked on migration and replication at the driver layer, mentored new contributors from the outreachy project, and provided bug fixes and reviews wherever possible. I'm a core reviewer and also a stable core. I have served as the release liaison for the last 2 cycles and run a weekly bug squashing meeting to help maintain our launchpad backog. We continue to face review bandwidth pressure. Emphasis should be placed on encouraging contributors to do code reviews and to recognize the efforts of those currently participating. There are several potential core reviewers and I feel that empowering the community is one way to increase engagement and the quality of our releases. 3rd party CI continues to be crucial resource for evaluating driver-specific changes. Maintaining good 3rd party CI for the drivers supported in Cinder by providing clear documentation and requirements will be a focus. Additionally, we have room for improvement in our backup and restore use case. My goal is to improve the entire backup/restore situation so that it can be leveraged by users in production environments. Our customers are using 3rd party backup products to backup workloads and volumes and I think this is one are where improvements here can directly benefit users. In summary, I would like to continue the efforts made by Rajat and continue to focus the areas most in need of attention for the 2024.2 cycle. I will do my best to make a positive contribution to both the technical and community aspects of Cinder during my time as PTL. Thank you for your consideration. -- Jon Bernard (jbernard)

1 0

Enabling external connectivity to VMs
by KK CHN 27 Feb '24

27 Feb '24

List, I am a bit confused with the OpenStack Networking configuration for Created virtual machines on a (Kolla-ansible all-in-one on an Ubuntu 22.04 VM ) openstack installation. I am able to create virtual machines through the dashboard as well as CLI. I have used the default network/cirros images created from init-runonce script. But Real time, how can I configure these new VMs to communicate with Base VM or external world by ping, ssh or make internet connection etc. Base VM is Ubuntu22.04 - with IPs ens192=>10.170.16.128(with Internet), second interface => ens224 => 10.220.0.51 and ens160 up and running without IP address for neutron_external_interface: All the VMs which I created getting IP addresses in series (10.0.0.X). But How to allocate Internet and make external world communication to and from these VMs ? Any hints/directions towards this much appreciated. Greetings, Krish If required more inputs here follows the additional information. (kolla_ansible_venv3) ubuntu@ubuntu-VM:~$ grep ^[^#] /etc/kolla/globals.yml --- workaround_ansible_issue_8743: yes config_strategy: "COPY_ALWAYS" kolla_base_distro: "ubuntu" openstack_release: "2023.1" kolla_internal_vip_address: "10.220.0.24" kolla_external_vip_address: "{{ kolla_internal_vip_address }}" kolla_container_engine: docker network_interface: "ens224" neutron_external_interface: "ens160" neutron_plugin_agent: "openvswitch" nova_compute_virt_type: "qemu" (kolla_ansible_venv3) ubuntu@ubuntu-VM:~$ Additional inputs: There are 3 virtual interfaces on the base Ubuntu22.04 Virtual Machine. ens192->10.170.16.128 with internet, ens224->10.220.0.51 , ens160 up without IP 1. First Interface contain IP address from My organizations DHCP lease from LAN(10.16.16.0) Which has Internet connectivity. ens192: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000 link/ether 00:0c:29:33:18:3c brd ff:ff:ff:ff:ff:ff altname enp11s0 inet 10.170.16.128/23 brd 10.170.17.255 scope global dynamic noprefixroute ens192 valid_lft 786963sec preferred_lft 786963sec inet6 fe80::cd1a:bc6c:b129:f91f/64 scope link noprefixroute valid_lft forever preferred_lft forever3 2. Interface I have 10.220.0.51 ens224: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000 link/ether 00:0c:29:33:18:46 brd ff:ff:ff:ff:ff:ff altname enp19s0 inet 10.220.0.51/24 brd 10.220.0.255 scope global noprefixroute ens224 valid_lft forever preferred_lft forever inet 10.220.0.24/32 scope global ens224 valid_lft forever preferred_lft forever inet6 fe80::a8b6:f094:c47:8282/64 scope link noprefixroute valid_lft forever preferred_lft forever 3. ens160 without any IP address which is up and running configured for Kolla neutron_external_interface: "ens160" ens160: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq master ovs-system state UP group default qlen 1000 link/ether 00:0c:29:33:18:32 brd ff:ff:ff:ff:ff:ff altname enp3s0 1. What networks need to be created additionally ? 2. Bridges and bonds required here or not ? 3. To enable these New VMs to talk to external world( ssh, ping, internet connection to all these VMs ) what all the steps need to be performed or I left out ?

2 1

Re: Use Openstack SDK for Swift
by gtema 27 Feb '24

27 Feb '24

> I didn't see a response to this email so I am re-sending. That's interesting because I responded. My Email client decided by default to respond to a single person dropping discuss, but response to you went out. Anyway, reposting my answer On Thursday, 22 February 2024 12:11:31 CET you wrote: > Hey > > this is a tricky thing explained by a single auth concept: in OpenStack when > you connect to the cloud as a regular user you request certain scope which > normally grants you access to the single domain or the project. Moreover > with this authorization you get list of endpoints you can communicate with > and for Swift this is a project bound endpoint that look somehow like > https:// swift.example.com/v1/AUTH_<project_id>. But the "AUTH_" part may > be actually different since it is part of the swift config. If you are an > admin the situation is a bit different > > In summary that means that you want to do something in the way the API was > not designed for. > > Solutions: > > generally first question you must address is the authorization: is your user > able to access containers of different projects or not. If not you have no > chance and you need to establish new connection with another credentials > When the answer is "yes" you have some options: > > - switch your authorization using > https://opendev.org/openstack/openstacksdk/ > src/branch/master/openstack/cloud/openstackcloud.py#L123. That gives you > new session for a different project that you can use in scope of a certain > project > > - use you single session for pure API calls. This has a disadvantage that > you have raw API communication and need to do lot of tricky things yourself > ``` > conn.get("../AUTH_<another_project_id>/") > ``` > > Unfortunately there is nothing else what could help you, but even with > alternative API wrappers you have the same issue. > > Hope this helps, > Artem > > On Tuesday, 20 February 2024 21:59:19 CET Albert Braden wrote: > > I'm writing a script that will use admin credentials to check user > > projects > > for containers. I'm already doing something similar for non-swift > > accounts, > > doing things like: > > > > cloud.network.networks(project_id=projectid): > > > > I read here about the Object SDK: > > > > https://docs.openstack.org/openstacksdk/latest/user/guides/object_store.ht > > ml > > > > And I tried this: > > > > cloud.object_store.containers(project_id=projectid): > > > > But I get an error: > > > > Invalid query params: project_id > > > > I read the Object SDK document here and it doesn't seem to say anything > > about how to specify a project: > > > > https://docs.openstack.org/openstacksdk/latest/user/proxies/object_store.h > > tm l > > > > But when I read the Swift API document, I see that I can use > > "/v1/{account}/{container}", so I'm guessing that there must be a way to > > do > > it via the SDK. > > > > https://docs.openstack.org/api-ref/object-store/#show-container-details-an > > d-> list-objects > > > > Is it possible to specify the project in an Object SDK call, or do I need > > to use the old python API for this?

2 1