- openstack-discuss - lists.openstack.org

neutron_server suspended in heath:starting
by Bursztynowski Dariusz 03 Nov '23

03 Nov '23

I've installed OpenStack (zed) on a Raspberry Pi cluster using kolla-ansible (version tagged for zed), and all containers are healthy except the neutron_server which is suspended in 'health: starting' state. Network related part of OpenStack seems not to work (snippet form running init-runonce): Configuring neutron. HttpException: 503: Server Error for url: http://192.168.1.60:9696/v2.0/routers, No server is available to handle this request.: 503 Service Unavailable. However, at least some commands not related to networking seem to work as expected (e.g., I can create an image which is reported by openstack image list as 'active'). There are four Raspberry Pi 4B in the cluster (2 x 4GB RAM and 2 x 8GB RAM). They run Debian 11 (bullseaye). Kolla-ansible has been used for the installation strictly following the guidelines form https://docs.openstack.org/kolla-ansible/zed/user/quickstart.html. Notably, I'm using a specific (weird?) configuration of networking on my Pis to mimic two network interfaces on each host as kolla-ansible expects. These are provided as interfaces of veth pairs (more details on that can be found in my report here: https://bugs.launchpad.net/neutron/+bug/2042598) Is it possible that this setup rises problems? I've already contacted OpenStack folks and they see no evident symptoms of neutron bug, suggesting a problem with deployment. Reported a bug to kolla-ansible here: https://bugs.launchpad.net/kolla-ansible/+bug/2042372, but do not know how fas they react there. I could try to install other versions but maybe there's a smarter ways to cope with the problem. Any suggestion will be welcome.

3 2

Provision of 100 HP ProLiant servers
by Leszek Szczepanowski 03 Nov '23

03 Nov '23

Hi all, I have the following use case: I need to install Openshift on 100 HP ProLiant servers, those are HP DL360 Gen10+. They are mounted in the racks, with factory settings. So, BIOS settings are default, as well as RBSU, RAID configs and ILO (ILO5). Is it possible to provision them using Openstack? Set a few parameters into ILO, into RBSU/BIOS, set up RAID controller disks (MegaRAID), and then put some ISO image and boot it? Or, maybe even better: create an Openshift cluster similar to how we do it with cloud providers? I have done this for years manually, via ILO WebUI + HTML5 console, but I'm actually tired of this. Can it be somehow automated? -- Leszek A. Szczepanowski twinsen(a)mspanc.net

4 3

[charms] Nominating Utkarsh Bhatt for charms-ceph-core
by Peter Sabaini 03 Nov '23

03 Nov '23

Hello all, I would like to nominate Utkarsh Bhatt for charms-ceph-core. Utkarsh has worked for the Ceph charms team for 1.5 years now and has provided many contributions to the charms, especially to the Ceph charms[0]. I know him as a diligent and enthusiastic contributor and hope you will support his nomination. Thanks, peter. [0] https://review.opendev.org/q/owner:utkarsh.bhatt%2540canonical.com

4 3

[neutron] Drivers meeting cancelled
by Brian Haley 03 Nov '23

03 Nov '23

Hello Neutrinos: Due to the lack of agenda, Friday's drivers meeting is cancelled. Have a nice weekend! -Brian

1 0

[ops] Seeking Advice on Service Deployment and Layout for an 8-Node Cluster
by Forrest Fuqua 02 Nov '23

02 Nov '23

Hi there, I'm currently exploring some strategies and seeking advice on the deployment and layout of services on an 8-node cluster. Each node is equipped with dual AMD EYPC 7702s, two 100G Links, and 2TB of RAM. Additionally, I've set up two 13TB NVME drives in a Ceph configuration for each node. At present, the setup involves one management node, while the others are dedicated to running Nova, Neutron, Glance, and Cinder. The workload, while fairly basic, involves handling substantial numbers. It's primarily used for one-off research projects and a few classes, serving around 60 end users. The most demanding scenario occurs during cybersecurity competitions when we need to deploy 3100 VMs at once. When this is spread out over a few days, it's manageable. However, there's a risk of overloading the system, leading to DiskIO issues and Database inconsistencies. Currently, I'm utilizing Kolla for deployment, and I've shared my configurations on this GitHub repository: Link to Configs<https://github.com/RIT-GCI-CyberRange/openstack-configs>. I'd greatly appreciate any insights or advice on optimizing this setup to handle the peak workload more efficiently and prevent potential performance issues. Thanks in advance for your help!

3 7

Call for Writers or OpenInfra Live Participants
by Kristin Barrientos 02 Nov '23

02 Nov '23

Hi everyone, Do you have a topic that you would like to discuss on OpenInfra Live or Superuser? We’d love to hear from you. Please reach out to me via email or IRC if you have any ideas and we can talk next steps. Thanks, Kristin Barrientos IRC: @kbarrientos:matrix.org Marketing Coordinator OpenInfra Foundation

1 0

[nova][ptg] 2024.1 Caracal PTG summary
by Sylvain Bauza 02 Nov '23

02 Nov '23

(sorry folks, was working on other urgencies hence the late summary) Again, as every cycle I say, that's a wrap ! Thanks to all the contributors that were around, we were more than 10 everyday, woohoo ! As a reminder (like every cycle ;-) ), please don't open the main etherpad link [1] if you use a Web browser with an automatic translation modification feature. If you do it, then it would internally translate all the phrases directly in the etherpad as if you were directly modifying the etherpad by yourself. Please also make sure you don't accidentally remove or modify lines. In order to prevent any misusage, this etherpad is a read-only copy of the main one : https://etherpad.opendev.org/p/r.02f889c0423aa279a3fa8e136becad10 Anyway, let's stop discussing about etherpad, and lemme provide you the summary : (oh and I could be wrong for explaining some stuff, no worries, just reply in here if so) (please also grab a coffee or whatever else, you may need it) ### Operator hour ### (read-only etherpad for this here https://etherpad.opendev.org/p/r.06c38397f2cdd81380649312f032b22f) So, we had around 4-5 operators for this operator hour, lovely. This is what we discussed : # flavors explosion for public clouds Basically, there are multiple usecases here : this is hard to discover flavor properties except by looking up the description, but this is also difficult to filter filters based on some properties since they can be different across clouds. What we agreed (as a start) : - Glance metadefs concept [2] could help to provide a catalog of flavor properties definition if operators were using it. - Nova could provide some new API microversion for filtering flavors based on a request of standard traits and/or resources (those queried within minimum and/or maximum bars). We don't know yet who could be the owner of such a specification, contributors welcome. - Another feature could be "Nova, tell me whether my flavor could have some candidates ?" but again, we need hands on deck. # Nova could get a better housekeeping system for orphaned resources # (I'm literally paraphrasing the topic name) OK, so, after discussing with the operator, we eventually found two use-cases : - as an operator, I'd like nova to magically adopt a compute node coming from elsewhere (say a region or another nova deployment). Long story short (lookup the etherpad for more details), the winning strategy that we could implement would be : #1 adopt a whole separate nova cell by using a specific nova-manage command that would create records on the target nova-api DB based from the records from the source api DB. The operator was happy with working on a POC and discussing with the nova community to make it a quickwin. #2 live-migrate my instances between cells. Well, that one is more difficult to implement, so we could punt this until #1 is implemented. - as an operator, I sometimes need to rebuild a compute OS. Well, on that one, we discussed it further more in terms of "project cleanup". Again, long story short, we went into explaining to use 'OSC project cleanup' [3] feature that'd do the job. # I want to be able to detach my root volume Well, that one wasn't really discussed, but the topic was filled in the etherpad. No real owner of such feature request, which absolutely requires a large community design discussion since resolving this isn't trivial (and we probably need to do some trade-offs in terms of guest support). Anyway, if you read those lines and you do care of such usecase, please understand that you need to associate resources and come by the community to bring them. I, as PTL, can help providing some mentoring and onboarding if needed. ### Cross-project meetings ### We had two cross-project sessions, one with Neutron and one with Cinder. # Neutron cross-project session We only discussed one topic, but a very important one : how Neutron would feel if Nova starts to require some optional Neutron API extensions ? :-) Eventually, what we agreed with Neutron is : - Nova will deprecate by Caracal the legacy codepath that doesn't use multiple port binding and provide a LOG.warning - Nova will delete this ^ in the E* release - Sean will provide a nova blueprint explaining which API extensions will be mandatory by Nova but also he will provide a neutron rfe bug report for that. - Eventually, we'll provide release notes and a ML thread for explaining all of this (we being Brian and me) # Cinder cross-project sessions We had two sessions, one in Wednesday, the other in Friday.. Please look at both [4] and [5] for knowing the summary for each of them. ### Bobcat retrospective and Caracal planning ### 17 blueprints were approved, 9 of them landed, one of them got reverted due to late issues discovered with OSC. 34 contributors this cycle, less than previously (mostly less on-off contributors). The list of untriaged bugs exploded, but some of our awesome contributors did a very late scrubbing, so now we're down to 30, huzzah. What we agreed at the PTG : - we gonna abandon the review-priority Gerrit label in favor of using an etherpad for both feature and bugs cycle tracking (we were already doing etherpad tracking for approved blueprints implementation patches) - that cycle tracking etherpad will provide a section for proposing patches, and another section only written by Nova PTL containing a scrubbed list of changes ready to review. Process to be further documented later (when our PTL eventually has time to do it) - PTL will review every approved spec to ensure that the testing requirements are reasonably defined Please also take care of our approved Caracal planning (I'll propose a release patch to add those deadlines into the main Caracal schedule page) : - Spec freeze at the second milestone - Spec review days on Tuesdays of the weeks R-21 and R-17 - Implementation review days on Tuesdays of the weeks R-20 and R-12 ### EOLing our unmaintained branches ### - bauzas will propose a patch for EOLing Ussuri - We'll defer until next cycle to see whether we drop Victoria and Wallaby ### SLURP cadence impact and curating Caracal ### - We are bumping our libvirt minimums early this cycle to Ubuntu Jammy versions - We will also update our next minimum versions later this cycle based on the numbers of the next Ubuntu 24.04 LTS release (once a beta build is available). - we'll forward port our Bobcat deprecation notices in the Caracal release notes - we end up merging all the required bits (mostly ORM changes) for preparing the SQLA 2.0 upgrade in a later release - we'll remove vmwareapi and hyperv support in Caracal - we agreed for the virt drivers removals to not provide an API microversion on the specific API endpoints and rather return a HTTP400 ### python tooling we could use ### - We agreed on using sphinx-lint, making the code changes having doc job to use it. - We agreed on testing codespell tool with the pep8 job and eventually either keep it or remove it depending on whether it finds many false positives - We agreed on reviewing the pyupgrade patches as a try with no clear promises. ### Planning new features ### - the Healthchecks spec is going to be reproposed this cycle, some implementation details about how to do RPC checks have to be identified - we agreed on exposing the value of the pinned AZ in GET /servers/<uuid>/details - we agreed on doing kind of soft-[anti-]affinity for AZs using instance groups (either thru new policies or some new concept of 'destination') - PCI device affinity for libvirt driver seems an effort to be resumed. Testing this is a concern but we'll discuss this in the spec - PCI devices notion of grouping (imagine two PCI devices for the same card, you'll get the idea) seems an interesting usecase, worth spec'ing. We could imagine the notion of a single placement unit being 'pci group' with one PCI device being the main object to track. The neighbouring aspect seems hard to implement so we reserve it for later. - Having console timeouts on token expiry seems a nice security feat, reviews welcome on the spec. - VGPU efforts will be huge this cycle. More to be discussed below in a specific section ### Caracal cleanups, bugs and follow-ups ### - we really want to change our default quota system to use unified limits soon but there are some upgrade concerns with flavors that contain placement resource classes that require ourselves to do further work with oslo.limits or Keystone in order to not break operators that chose to use the unified limits driver. We could also add some checking in the nova-manage migration tool that would emit a warning if resource classes were used. - our config with images_types_backend, use_cow_images and force_raw_images is confusing and errorprone. We agreed on fixing some migration bug as a backportable way, but also deprecating use_cow_images and force_raw_images while providing meaningful config options instead. - we're mixing IP addresses and URLs usage in our configs, we could be more consistent. What we agreed on that is that a new migration_inbound_addr config sounds cool, defaulting to the my_ip config value but being able to provide a FQDN or hostname too ### The VGPU efforts ### - we're actively working on providing upstream testing for mdevs, thanks to the mtty kernel samples framework. mtty live-migration is currently an ongoing patch, but someone proposed to backport this into a custom kernel built on purpose for our usage. Either way, mtty testing will be done as much as possible in nova-next but we could create a new periodic/experimental target for targeting new features like live-migration. - we plan to do mdev live-migration this cycle. There are lots of limitations that we need to document but this sounds doable. We could add some pre-live-migration check about mdev types and this would require some object change. We may also need to persist the mdev/allocation relationship but that will be discussed in the spec - nvidia is going to drop vfio-mdev for their SR-IOV GPUs and use a specifc vfio-pci variant driver. This is a very premature task and our caracal efforts will consist in identifying the necessary nova adaptations. One is already identified (adding a managed flag) and we agreed on a solution. - we're gonna change how nova creates the mediated devices by rather defining a libvirt nodedevice which will also facilitate operator maintenance (they could just mdevctl or define udev rules) - we agreed on deprecating the config default that allows you to only specify mdev types without specifying the addresses. - SRIOV GPUs currently leak some unused VFs into Placement, leading to capacity issues [6]. We could address that problem by either making the device's explicit definition mandatory or adding a new config option that would specify the number of VFs to use per type. - some other bugs were discussed with an operator who was present. We agreed on doing further testing to clearly identify problems, and some patches are already up for reviews (like the evacuate case or the multi-create limitation) but also require some rebase. Anyway, I guess I'm done now. Thanks for having read until that point and I hope your coffee (or tea) was good. HTH and thanks, -Sylvain (on behalf of the whole Nova community) [1] https://etherpad.opendev.org/p/nova-caracal-pt g (please remove the empty char between 'pt' and 'g') [2] https://docs.openstack.org/glance/latest/user/metadefs-concepts.html [3] https://docs.openstack.org/python-openstackclient/latest/cli/command-object… [4] https://wiki.openstack.org/wiki/CinderCaracalPTGSummary#Cross_project_with_… [5] https://wiki.openstack.org/wiki/CinderCaracalPTGSummary#Nova_Cinder_cross_p… [6] https://bugs.launchpad.net/nova/+bug/2041519

1 0

[sunbeam][ptg] vPTG summary
by James Page 02 Nov '23

02 Nov '23

HI All The Sunbeam project participated in the vPTG last week; here is a short summary of the topics that we discussed and plans added for the Caracal cycle. # Full Cycle Retro ## Celebrations - Significant reduction in both the time taken to bootstrap and resize a deployed cluster. - Switch from Antelope to Bobcat in the charms was smooth and relatively quick. - Many new plugins delivered during the cycle, enabling lots more of the OpenStack ecosystem - telemetry, magnum, heat, barbican and designate. - Refactoring of ops-sunbeam to support both k8s and machine charms was a success - Documentation starting to look great and developing lots of depth - Lots of contributions to the open source tools we depend on including snapd, Juju (inc terraform provider), microk8s, traefik and prometheus charms. ## Learnings - Switching to Bobcat for snaps and terraform plans took a bit longer than it should have done - most likely down to this being a largely new process for the team. - Some impact on velocity due to bugs in dependencies outside of OpenStack - Release process needs to be better documented so more of the core team can execute it. - Plugin full functional test coverage still needs some work. ## Actions - Implementation of automated functional testing for plugins - Full documentation and improved automation of the release process # Debugging Problems ## Overview We had quite an extended conversation about how we help users debug both bugs and deployment specific issues in Sunbeam; the general observability feature we're working on will help in this area once a cloud is deployed but we need some extra features to help users before this stage of the deployment is reached. sunbeam inspect is a bit lightweight still; can supplement the information gathered with more extensive information collected from microk8s so we need to document this into the bug reporting guidance. ## Actions - Improve docs topic 'Inspecting the cluster' to include more detail of collecting debug information for reporting bugs - Update the guidance for reporting bugs in Launchpad projects to help users report better more qualified bugs # Monorepo for Sunbeam Charms ## Overview With our current approach of 1:1 mapping between charms and repositories, we're going to have a similar challenge to the OpenStack Charms project when we need to effect a change across the entire charmset (bumping to a new release, updating a dependency or alike). This generates a large number of reviews, and associated costs of testing, which then need to be managed through to landing by one of the core team. Liam proposed that we look at moving to a single mono-repo for all of the charms in the project, reducing the overhead of 'batch' type operations and making it generally easier to manage this part of Sunbeam. We'd also address integrated testing of a change that impacts multiple charms without having to model cross-repo dependencies. The resource required to complete a function test would increase in size, but reduce hugely in number when compared to our current approach. We'll need to figure out tooling to determine which charms are impacted by a specific change so that we only build, test and publish charms that need to be changed, rather than the entire charmset. Some projects in the wider charm ecosystem are using this approach to good effect. ## Actions - Liam to take this idea forward early during the Caracal cycle - will need some coordination and advice from the infra team (see next topic). # Integration testing ## Overview Although the Sunbeam charms do include functional testing, it's not a complete integration test of a full Sunbeam deployment. We'd like to achieve this goal but this will require test instances with more memory than we typically get right now. We discussed that this felt linked to the idea of the mono-repo for charms as it feels unreasonable to ask for increased instance sizes for and increasing set of repositories when we could achieve better resource usage efficiency with a mono-repo. We agreed to fold this into the mono-repo work early this cycle. # Multi-node testing ## Overview All of our multi-node testing is currently not executed as part of the project CI workflows - we really need to test this part of Sunbeam as part of our gating for changes. We discussed a few ideas in this area including use of LXD containers on a big instance type or use of the 3rd party CI provided by Canonical that currently supports the OpenStack Charms project. Both might present viable routes forward and will be considered during the coming cycle. # Managing OCI artefacts references in charms ## Overview The OCI resources used by the Sunbeam K8S charms are managed within a different project; we need to be aware of when new OCI's are published so we can feed these updates into the charms and publish new charm revisions and associated OCI resources. There is some prior art on github modelled as actions to support this type of workflow but we need something that will work across that boundary into the opendev gerrit infrastructure. ## Actions - Reach out to the opendev infrastructure team as to whether there is any prior work on this type of function. # Single Container Multiple Services ## Overview We've spent quite a bit of time this cycle reducing the physical footprint of the OCI resources we use for Sunbeam. We discussed whether this might be taken further by reducing the number of containers we run in each charm - this is made possible by the fact that the OCI entry point is managed using Pebble, which supports multiple service definitions so we can run multiple daemons within the same container. We agreed to prototype this in one of the more complicated charms to see how much benefit might be achieved by taking this approach. ## Actions - POC probably using the nova-k8s charm to assess benefits. We had around 6 active participants during our session with a few others dropping in to listen in on the conversation from time to time - thankyou to all of those who contributed to our discussions and I look forward to the next (v)PTG. Cheers James Sunbeam PTL

1 0

[kolla-ansible][16.1.0] Could not load namespace_haproxy_for_cell
by Satish Patel 02 Nov '23

02 Nov '23

Folks, I am deploying new cloud and planning to use 2023.1 with kolla-ansible tag 16.1.0 but I hit this error and after google I found this reference bug: https://bugs.launchpad.net/kolla-ansible/+bug/2039573 Very curious about the workaround here. How does it pass CI-CD if it's really a bug?

3 3

[nova][compute] HugePages question
by Satish Patel 01 Nov '23

01 Nov '23

Folks, What are the advantages and disadvantages to enable HugePages on compute nodes? I know that HugePages improve performance of VMs because of TLB lookup etc. But if it does improve in most cases then why people why default not turning on HugePage for deployment? Just trying to find out what could go wrong if I turn on by default HugePage for all deployment?

2 2