May 2024 - openstack-discuss - lists.openstack.org

launch vm and its LVM cinder volumes on same host
by Morteza FarkhondePour 15 May '24

15 May '24

0 <https://serverfault.com/posts/1159435/timeline> I have an OpenStack *Antelope* version installed via Kolla-Ansible, and I have several compute hosts with LVM volume types. When I create an instance using an LVM volume, I expect both the volume and the instance create on the same host. However, they end up on different hosts. For example, the volume is located on d13-compute43@localA#localA, whereas the instance is on d13-compute51

1 0

[kolla] Weekly meeting today (15th May 2024) cancelled
by Michał Nasiadka 14 May '24

14 May '24

Hello Koalas, Weekly meeting today is cancelled due to key participants (including me) not being able to attend. Let’s meet next week (22nd May 2024) - see you then! Michal

1 0

OpenInfra Summit Asia'24 CFP closing in two weeks
by Jeremy Stanley 14 May '24

14 May '24

The OpenInfra Summit Asia'24 is the first community-led Summit[0], hosted by the OpenInfra Korea User Group in collaboration with other OpenInfra user groups throughout Asia. The CFP is currently open and accepting applications until May 29, 2024, so there's only two weeks left to get your proposals submitted[1]. Keep in mind that Summit talks are often professionally recorded and re-watched by a much larger global audience, having a broad impact for years after the event concludes, so your reach can be far greater than just the people who happen to be in the room at that moment. This time, the Summit will be co-located with the OCP (opencompute.org) Regional Summit APAC and will showcase the latest advancements in open source infrastructure software and hardware. If you want to present but are struggling to to come up with ideas, "beginner" content is always a crowd pleaser (setting up your first environment, tips for new contributors), but this could also be of particular interest for the growing local community in Korea. Since this event is occurring alongside one for the Open Compute Project, hardware-related topics might serve to draw in participants from their community and help grow our own through the sort of collaboration we're best known for. Upgrade stories and talks about reducing complexity are also something our users are especially keen to hear about. For that matter, talks about Networking (including 5G telephony) or integration with adjacent community tech (Ceph, Kubernetes, OVN) would be great to show how OpenStack is really an integral piece of the larger open source ecosystem. If you want to attend and your employer cannot fund your travel, the OpenInfra Foundation is continuing the OpenInfra Travel Support program. You can apply for travel support[1] until May 29. If you have any questions or want help (even if it's just to bounce ideas off me for a talk you're thinking about proposing), feel free to reply or find me (fungi) in the #openstack channel on the OFTC IRC network. [0] Register for the OpenInfra Summit Asia: https://2024.openinfraasia.org/ [1] Submit to the CFP: https://openinfrafoundation.formstack.com/forms/openinfra_asia_summit_2024 [2] Apply for the OpenInfra Travel Support Program: https://openinfrafoundation.formstack.com/forms/openinfra_tsp -- Jeremy Stanley

1 0

Boot an uefi volume that migrated from VMware ESXI
by luckydog xf 14 May '24

14 May '24

Hi, list, I migrated a volume from VMware ESXI VMDK, and originally on VMware platform it's EFI boot mode and secure boot enabled. I cloned the entire VMDK by clonzilla and restored on OpenStack as a volume. I referred to this link https://docs.openstack.org/nova/latest/admin/secure-boot.html and uploaded that volume as an image. Then set properties. Finally, I launched an instance with a target image, however, it cannot boot up with a blank and dark screen. What's the right way to implement this? Thanks.

2 11

[nova] Our first spec review day for Dalmatian
by Sylvain Bauza 14 May '24

14 May '24

Hey folks, As agreed during the PTG, we'll have two spec review days this cycle. Next Tuesday, ie. May 14 2024, we'll have the first one. Please make sure you can upload your spec before this and please be around on IRC on Tuesday (if you can) so we could discuss your spec change if we need. Thanks, -Sylvain

1 1

[openstack][kolla-ansible][keystone]upgrade from yoga to zed error keystone database
by Nguyễn Hữu Khôi 13 May '24

13 May '24

Hello. I upgraded from yoga to zed but i encounter this error as a keystone log. https://www.editpad.org/?edit-id=CNujHK962K70e93ae8 I tried to remove this table but this error still appears with other tables. My env: ubuntu22.04 openstack with kolla ansible as the deployment tool Thank you. Regards Nguyen Huu Khoi

2 4

[neutron] Bug deputy report for week of May 6th
by Miguel Lavalle 13 May '24

13 May '24

Hi, I was the Neutron bugs deputy last week. Following is a brief summary of the reported bugs: Critical bugs ----------------- None High bugs -------------- https://bugs.launchpad.net/neutron/+bug/2065094 OVN revision number is bumped on noop port update. Assignee Jakub Libosvar. Released fix: https://review.opendev.org/c/openstack/neutron/+/918492 https://bugs.launchpad.net/neutron/+bug/2065459 ovn-octvia-provider: RuntimeError: cannot join current thread. In progress. Assignee Terry Wilson. Proposed fix: https://review.opendev.org/c/openstack/ovn-octavia-provider/+/919229 https://bugs.launchpad.net/neutron/+bug/2065460 ovn-octavia-provider results in a new OVSDB connection for every Octavia API request. In progress. Assignee Terry Wilson. Proposed fix: https://review.opendev.org/c/openstack/ovn-octavia-provider/+/919229 https://bugs.launchpad.net/neutron/+bug/2065566 [ovn-octavia-provider] Return DELETED on member_delete for non-existing members. In progress. Assignee Fernando Royo/ Proposed fix: https://review.opendev.org/c/openstack/ovn-octavia-provider/+/919482 Medium bugs ------------------ * https://bugs.launchpad.net/neutron/+bug/2065145 AttributeError: 'MetadataDriver' object has no attribute 'monitors' in L3 agent logs, Assignee Brian Haley. Proposed fix: https://review.opendev.org/c/openstack/neutron/+/918674 Low bugs ------------- None Wishlist bugs ------------------ https://bugs.launchpad.net/neutron/+bug/2065198. Allow ml2 MechanismDrivers to start own rpc listeners. Gathering more input from submitter at the moment

1 0

Nova filters
by Gk Gk 13 May '24

13 May '24

Hi, In the Antelope version nova scheduler, is there an option for filtering hypervisors based on available disk space ? I cant find the option "DiskFilter' in the latest documentation for antelope. How to achieve this ? Thanks G.K

2 4

[DNS] RFH: Something is answering DNS PTR queries with UDP and not EDNS
by Philipp Hahn 13 May '24

13 May '24

Hello, I'm new to this list and relatively new to OpenStack, so please be patient with me. Our infrastructure provider is hosting OpenStack for us and I observed the following irritating behavior: If my Linux VM does a reverse-IP-lookup (PTR) something™ seems to intercept those PTR queries on their way from the VM to the *configured* DNS server and answers them itself, returning the name of the VM inside some pre-configured domain: > # dig +nocmd +nocomments +nostats +noedns -x 10.201.0.211 > ;211.0.201.10.in-addr.arpa. IN PTR > 211.0.201.10.in-addr.arpa. 3600 IN PTR jenkins-ucs-5-0-ucs-5-0-7-autotestjoinopenstack-sambaversion-no.a.customers.regiocloud.tech. > > # getent hosts 10.201.0.211 > 10.201.0.211 jenkins-ucs-5-0-ucs-5-0-7-autotestjoinopenstack-sambaversion-no.a.customers.regiocloud.tech > > # python3 -c 'import DNS;req=DNS.DnsRequest("211.0.201.10.in-addr.arpa.",qtype="PTR",server=["10.201.0.211"], protocol="udp");res=req.req();print(res.answers[0]["data"])' > jenkins-ucs-5-0-ucs-5-0-7-autotestjoinopenstack-sambaversion-no.a.customers.regiocloud.tech > > # python3 -c 'import dns.resolver;r=dns.resolver.Resolver(configure=False);r.nameservers=["10.201.0.211"];r.edns=False;a=r.query("211.0.201.10.in-addr.arpa.","PTR",tcp=False);print(list(a))' > [<DNS IN PTR rdata: jenkins-ucs-5-0-ucs-5-0-7-autotestjoinopenstack-sambaversion-no.a.customers.regiocloud.tech.>] This only happens when UDP is used an EDNS is disabled: As soon as I use TCP or force EDNS my queries arrives at the configured DNS server and I get back the expected answer: > # dig +nocmd +nocomments +nostats +edns -x 10.201.0.211 > ;211.0.201.10.in-addr.arpa. IN PTR > 211.0.201.10.in-addr.arpa. 172800 IN PTR master096.autotest096.test. > 201.10.in-addr.arpa. 10800 IN NS master096.autotest096.test. > master096.autotest096.test. 80600 IN A 10.201.0.211 > > # dig +nocmd +nocomments +nostats +edns +tcp -p 53 -x 10.201.0.211 > ;211.0.201.10.in-addr.arpa. IN PTR > 211.0.201.10.in-addr.arpa. 172800 IN PTR master096.autotest096.test. > 201.10.in-addr.arpa. 10800 IN NS master096.autotest096.test. > master096.autotest096.test. 80600 IN A 10.201.0.211 > > # dig +nocmd +nocomments +nostats +noedns +tcp -p 53 -x 10.201.0.211 > ;211.0.201.10.in-addr.arpa. IN PTR > 211.0.201.10.in-addr.arpa. 172800 IN PTR master096.autotest096.test. > 201.10.in-addr.arpa. 10800 IN NS master096.autotest096.test. > master096.autotest096.test. 80600 IN A 10.201.0.211 > > # python3 -c 'import DNS;req=DNS.DnsRequest("211.0.201.10.in-addr.arpa.",qtype="PTR",server=["10.201.0.211"], protocol="tcp");res=req.req();print(res.answers[0]["data"])' > master096.autotest096.test > > # python3 -c 'import dns.resolver;r=dns.resolver.Resolver(configure=False);r.nameservers=["10.201.0.211"];r.edns=False;a=r.query("211.0.201.10.in-addr.arpa.","PTR",tcp=True);print(list(a))' > [<DNS IN PTR rdata: master096.autotest096.test.>] > > # python3 -c 'import dns.resolver;r=dns.resolver.Resolver(configure=False);r.nameservers=["10.201.0.211"];r.edns=True;a=r.query("211.0.201.10.in-addr.arpa.","PTR",tcp=True);print(list(a))' > [<DNS IN PTR rdata: master096.autotest096.test.>] > > # python3 -c 'import dns.resolver;r=dns.resolver.Resolver(configure=False);r.nameservers=["10.201.0.211"];r.edns=True;a=r.query("211.0.201.10.in-addr.arpa.","PTR",tcp=False);print(list(a))' > [<DNS IN PTR rdata: master096.autotest096.test.>] I already talked to out provider and spend multiple hours searching the internet or even having a look at the source code of Neutron myself, but did not find any clue so far. Can someone please enlighten me and tell me, how I (or our provider) can disable that "UDP DNS intercept behavior" as that breaks our current use-case, especially as I seldomly can configure our applications to (not) use UDP / EDNS, so depending on which resolver library they use they get different answers to the same query. Thank you in advance. Philipp Hahn -- Philipp Hahn Open Source Software Engineer Univention GmbH Mary-Somerville-Str. 1 28359 Bremen Germany | Deutschland Phone: +49 (0)421 22232-0 | E-Mail: info(a)univention.de https://www.univention.de | https://www.univention.com Managing Directors: Peter H. Ganten, Stefan Gohmann Local court: Amtsgericht Bremen HRB 20755 | Ust-ID: DE220051310 If you are not the intended recipient of this mail, please contact the sender and delete this message. Any unauthorized copying of this message or unauthorized distribution of the information contained herein is prohibited.Information on the processing of your personal data can be found here: https://www.univention.com/privacy-statement

2 1

Re: CAD applications on OpenStack
by Francesco Di Nucci 13 May '24

13 May '24

You're welcome have you also thought about having a chat with CAD vendor's team? Maybe they encountered it as a problem, as it's more about the CAD and less about having Windows VMs on OpenStack By the way, when a conversation is in a list, reply to all to include the list ;) Best regards Francesco Di Nucci On 13/05/24 12:20, Arne Wiebalck wrote: > Thanks, Francesco! > > Licenses are a concern and we're looking at that. > > What we're not really clear on if this approach works latency- / concurrency wise. > We will probably simply try :) > > Thanks again! > Arne > > -- > Dr. Arne Wiebalck > Group Leader, Compute and Devices Group > CERN IT > > ________________________________________ > From: Francesco Di Nucci <francesco.dinucci(a)na.infn.it> > Sent: Thursday, 9 May 2024 08:12 > To: Arne Wiebalck > Subject: Re: CAD applications on OpenStack > > Hi, > > former CAD user here - first of all I'd advice to clear with your legal > department if your license allows you to use the CAD in VMs. > > AFAIK CATIA runs on Windows, so containerization is not possible unless > you manage to get the CAD running in Wine, and you should configure PCI > passthrough > > (https://docs.openstack.org/nova/2024.1/admin/pci-passthrough.html) to > use a graphic card in the VMs. > > Regards > > Francesco Di Nucci > > On 05/05/24 12:04, Arne Wiebalck wrote: >> Dear all, >> >> We're looking into integrating the use case of CAD applications, such as CATIA V5/6 or CATIA 3DEXPERIENCE, into our OpenStack deployment, with the goals of increased resource sharing >> and to reduce the park of individual workstations we have for this at the moment. >> >> If anyone has experience running such applications on top of OpenStack (on individual VMs >> with GPU access, on bare metal servers with concurrent sessions, maybe even containerised ...) >> and would like to share what they have done, it'd be great if you got in touch with me. >> >> Thanks! >> Arne >> >> -- >> Arne Wiebalck >> CERN IT >> >> > >

1 0

Re: Nova filters
by Gk Gk 13 May '24

13 May '24

Hi, In the Antelope version nova scheduler, is there an option for filtering hypervisors based on available disk space ? I cant find the option "DiskFilter' in the latest documentation for antelope. How to achieve this ? Thanks G.K

1 0

Rabbit mq connections keep on going up, why channels and connections not closing?
by Shrey Chauhan 13 May '24

13 May '24

Hi all, I am observing a behaviour which I am not sure is right or wrong, or something I should be concerned about? Currently I am using openstack xena with rabbit-mq message broker. rabbit mq version is 3.8.3 amqp version 5.0.8 I have installed rabbit management plugin on my cluster, and I observe whenever I perform any task like a resize. My connection and channel counts go up, and with every operation it goes up. I am worried that at some point I will hit some kind of connection threshold. I just want to understand if services like nova, cinder reuse an open rabbit mq connection or reopen for every operation, and if they open msg queue connection for every operation do they close the connection once the action is done Am I missing configuration in my setup of rabbit mq with openstack? Please help me understand this scenario. Thanks for help in advance. Thanks, Shrey Chauhan

1 0

[ops] Inquiry Regarding Addition of DH Groups in OpenStack VPNaaS
by Shubham Yadav 13 May '24

13 May '24

Hello , I hope this message finds you well. I am reaching out to inquire about the possibility of adding additional DH (Diffie-Hellman) groups to the VPNaaS (VPN as a Service) project in OpenStack. Currently, our setup involves using DH groups 14, 16, 20, and 21 on Fortigate routers. However, the available DH groups at the OpenStack Web UI for creating IPSEC policies are limited to groups 2, 5, and 14. Could you please provide insights into whether there are plans to introduce more DH groups into the VPNaaS project? This addition would greatly benefit our configurations and improve compatibility with our existing infrastructure. Thank you in advance for any information or guidance you can provide on this matter. Best regards, Shubham

2 1

[oslo][release] Changing back release model of tooz ?
by Takashi Kajinami 13 May '24

13 May '24

Hello, As you may know, tooz was moved to the independent release model[1] after wallaby release because of its stable code base at that time. However, we introduced several features/improvements in tooz during a few cycles and I'm wondering if we can revisit its status now. One actual problem I'm seeing now is that, we recently got a bug report[2] about the redis sentinel support which we added during 2024.1 cycle but we can't backport the required bug fix to 2024.1 because of no stable branch. Given the situation, I'm wondering if it makes sense (and is possible) to move tooz back to the cycle-with-intermediary model, and create 2024.1 branch from 6.1.0 (the last release during 2024.1 cycle) now, so that we can maintain 2024.1 releases with future bug fixes brackported. [1] https://review.opendev.org/c/openstack/releases/+/788230 [2] https://bugs.launchpad.net/python-tooz/+bug/2063857 -- Takashi Kajinami irc: tkajinam github: https://github.com/kajinamit launchpad: https://launchpad.net/~kajinamit

4 5

[tc][all] OpenStack Technical Committee Weekly Summary and Meeting Agenda (2024.2/R-20)
by Goutham Pacha Ravi 12 May '24

12 May '24

Hello Stackers, We're officially in milestone-1 week for the 2024.2 (Dalmatian) release cycle [1]. Teams are actively planning design specification freezes, hackathons, and bug squashes. I'm delighted to announce that three Outreachy interns have been funded to work with the OpenStack community [2]. Please join me in welcoming Mohd Ariq (tspyderboy), Okeke Christian Chima (cokeke), and Abby Nduta (AbbyNduta). Mohd Ariq and Christian will be working with members of the OpenStack Manila project team, while Abby will be working with members of the OpenStack Horizon project team. Congratulations, interns! These internships begins at the end of May and will conclude at the end of August. There have been a few governance changes in the past week that I'd like to brief you about: The "Consistent and Secure Default RBAC" goal has new milestone requirements. All projects are expected to transition oslo.policy's "enforce_scope" and "enforce_new_defaults" options to True during the 2024.2 release cycle. These options will be removed from oslo.policy during the 2025.1 release cycle. [3] openstack/kuryr and openstack/kuryr-libnetwork are now under the maintainership of the openstack/zun ("Container as a service") team. The TC has adopted a house rule to mark projects that have identified no leadership as "inactive" projects [4]. We've merged slaweq's summary of the 2023 OpenStack User Survey [5]. Thanks, slaweq, and to all of you who took the survey. === Weekly Meeting === Last week's TC meeting was held simultaneously on Zoom (a recording is on YouTube [6]) and IRC on May 7th, 2024. The meeting was logged to eavesdrop [7]. Over the years, we've come to realize that monitoring project activity through objective tools wasn't representing the picture accurately. Several mature/feature-complete projects will seemingly be "inactive," and what we're interested in assessing is whether there are active maintainers tackling security and project infrastructure concerns. Artem Goncharov (gtema) will write up a proposal to tag inactive projects more prominently. We discussed a proposal to change the Distributed Project Leadership model to require DPL liaisons to re-assert their interest at the beginning of each cycle. The proposal also suggests appointing liaisons from the TC to monitor projects that have this leadership model. TC Members would be welcome to nominate themselves as liaisons for projects following DPL if this proposal is accepted. We took some AIs to continue the cleanup of maintainers on OpenStack's packages on PyPI - an ongoing community-wide effort [8]. We discussed the open goal proposal to replace eventlet. The Technical Committee wants to ensure that we're not advocating for a one-size-fits-all solution to the problem of threading in OpenStack projects. Hence, we'd prefer not to merge the proposal to replace eventlet with asyncio. Instead, we want to encourage projects to independently explore solutions that make sense to their individual architecture - which could be asyncio, native threads, or any other threading paradigm/library. We are supportive of ensuring there's no imposition of any specific library or threading mechanism imposed via oslo libraries. JayF will be summarizing this view to this mailing list soon. The meeting concluded with an open discussion on DevStack setup time and an outdated (but relevant) proposal related to OSC as a service on DevStack. Tomorrow's meeting is back on OFTC/IRC, and the call details and agenda are at [9]. We look forward to seeing you there. The TC is always honored to chat with you, so please join us, and feel free to add anything to the agenda that you'd like to chat with us about! === Goal Proposals === As stated above, this goal proposal needs your weigh-in: - Migrate eventlet usages to asyncio (https://review.opendev.org/c/openstack/governance/+/902585) === Governance Proposals === ==== Merged ==== - Remove ref of bylaw Appendix 4 | https://review.opendev.org/c/openstack/governance/+/916942 - Add timeline to remove enforce_scope in RBAC goal | https://review.opendev.org/c/openstack/governance/+/915179 - Move openstack/kuryr and openstack/kuryr-libnetwork to Zun | https://review.opendev.org/c/openstack/governance/+/917517 - 2023 User Survey TC Question Analysis | https://review.opendev.org/c/openstack/governance/+/900516 - Add option to move leaderless project to Inactive status | https://review.opendev.org/c/openstack/governance/+/914726 ==== Open for review/discussion ==== - Update to include docs and miscellaneuos repos for AC status | https://review.opendev.org/c/openstack/governance/+/915021 - Add DPL model & liaison reset policy | https://review.opendev.org/c/openstack/governance/+/916833 Thank you very much for reading! On behalf of the OpenStack TC, Goutham Pacha Ravi (gouthamr) OpenStack TC Chair [1] https://releases.openstack.org/dalmatian/schedule.html [2] https://www.outreachy.org/alums/2024-05/#:~:text=Openstack [3] https://governance.openstack.org/tc/goals/selected/consistent-and-secure-rb… [4] https://governance.openstack.org/tc/reference/emerging-technology-and-inact… [5] https://governance.openstack.org/tc/user_survey/analysis-2023.html [6] https://youtu.be/-WZT7PC3sig [7] https://meetings.opendev.org/meetings/tc/2024/tc.2024-05-07-18.00.log.html [8] https://etherpad.opendev.org/p/openstack-pypi-maintainers-cleanup [9] https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Next_Meeting

1 0

openstack boot an instance from an IDE volume migrated from VMware
by luckydog xf 10 May '24

10 May '24

hi, list, I'm going to boot an instance from an existing volume, and it is a Windows 2016 migrated from VMware. So the bus must be set to 'ide' mode, otherwise it fails to boot up. Let's assume this volume UUID is xyz and it's bootable. I cloned it by clonezilla from VMware ESXI. since virtio drive wasn't installed in Windows, so it must be set to ide. So what's the right way to boot it from this volume and set it to IDE? I have tried lots of ways, like --- openstack server create test-windows --flavor 4 --nic net-id=7b57ee99-136f-4aa7-b3fe-a4958ec8c644 --block-device id=xxxx:disk_bus=ide and ... --block-device-mapping. But unfortunately, none of them work. And there are too many outdated docs like disk_bus=ide and bus=ide. Not sure which one works. This is fairly frustrating. I'm still struggling with this issue and need your help. Thanks a lot.

3 7

[tc][all][ec2-api] Retiring the ec2-api project
by Ghanshyam Mann 10 May '24

10 May '24

Hello everyone, As you might know, the ec2-api project is not active and was marked 'inactive' since 2024.1 cycle[1]. In this cycle also, this project is leaderless[2] and no maintainer stepped up to maintain it. As discussed in the Leaderless project etherpad[2], the next step is to retire the project. I'm sending this email to propose the ec2-api retirement and based on the response, I will proceed next. [1] https://governance.openstack.org/tc/reference/emerging-technology-and-inact… [2] https://etherpad.opendev.org/p/2024.2-leaderless#L72 -gmann

1 1

[tc][all][senlin] Retiring the Senlin project
by Ghanshyam Mann 10 May '24

10 May '24

Hello everyone, As you might know, the Senlin project is not active and was marked 'inactive' since 2024.1 cycle[1]. In this cycle also, this project is leaderless[2] and no maintainer stepped up to maintain it. As discussed in the Leaderless project etherpad[2], the next step is to retire the project. I'm sending this email to propose the Senlin retirement and based on the response, I will proceed next. [1] https://governance.openstack.org/tc/reference/emerging-technology-and-inact… [2] https://etherpad.opendev.org/p/2024.2-leaderless#L72 -gmann

1 1

[tc][all][solum] Retiring the Solum project
by Ghanshyam Mann 10 May '24

10 May '24

Hello everyone, As you might know, the Solum project is not active and was marked 'inactive' since 2024.1 cycle[1]. In this cycle also, this project is leaderless[2] and no maintainer stepped up to maintain it. As discussed in the Leaderless project etherpad[2], the next step is to retire the project. I'm sending this email to propose the Solum retirement and based on the response, I will proceed next. [1] https://governance.openstack.org/tc/reference/emerging-technology-and-inact… [2] https://etherpad.opendev.org/p/2024.2-leaderless#L72 -gmann

1 1

[tc][all][murano] Retiring the Murano project
by Ghanshyam Mann 10 May '24

10 May '24

Hello everyone, As you might know, the Murano project is not active and was marked 'inactive' since 2024.1 cycle[1]. Due to a critical bug in Murano, the TC needs to pass a resolution to stop its 2024.1 release[2]. In this cycle also, this project is leaderless[3] and no maintainer stepped up to maintain it. As discussed in the Leaderless project etherpad[3], the next step is to retire the project. I'm sending this email to propose the Murano retirement and based on the response, I will proceed next. [1] https://governance.openstack.org/tc/reference/emerging-technology-and-inact… [2] https://governance.openstack.org/tc/resolutions/20240227-remove-murano-from… [3] https://etherpad.opendev.org/p/2024.2-leaderless#L72 -gmann

1 1

[magnum][magnum-cluster-api] Scalling Master Node
by pahrialtkj＠gmail.com 10 May '24

10 May '24

Hello all, I created a k8s cluster with 1 master and 1 worker node, I want to increase the master node to 2 or 3, I tried to do it with cluster resize/update and nodegroup create seems to be not supported and does not work, it works for the worker node, has anyone tried scaling the master node?

3 3

Re: How to check the free RAM or CPU or Disk in openstack python module
by Lajos Katona 10 May '24

10 May '24

Hi, No I meant something like this: $ openstack resource provider list +--------------------------------------+-------+------------+--------------------------------------+----------------------+ | uuid | name | generation | root_provider_uuid | parent_provider_uuid | +--------------------------------------+-------+------------+--------------------------------------+----------------------+ | 0f916a23-403f-43c3-ac57-e109bc371dba | jammy | 2 | 0f916a23-403f-43c3-ac57-e109bc371dba | None | +--------------------------------------+-------+------------+--------------------------------------+----------------------+ $ openstack resource provider inventory list 0f916a23-403f-43c3-ac57-e109bc371dba +----------------+------------------+----------+----------+----------+-----------+-------+------+ | resource_class | allocation_ratio | min_unit | max_unit | reserved | step_size | total | used | +----------------+------------------+----------+----------+----------+-----------+-------+------+ | VCPU | 4.0 | 1 | 6 | 0 | 1 | 6 | 0 | | MEMORY_MB | 1.0 | 1 | 7936 | 512 | 1 | 7936 | 0 | | DISK_GB | 1.0 | 1 | 28 | 0 | 1 | 28 | 0 | +----------------+------------------+----------+----------+----------+-----------+-------+------+ $ openstack resource provider inventory list 0f916a23-403f-43c3-ac57-e109bc371dba +----------------+------------------+----------+----------+----------+-----------+-------+------+ | resource_class | allocation_ratio | min_unit | max_unit | reserved | step_size | total | used | +----------------+------------------+----------+----------+----------+-----------+-------+------+ | VCPU | 4.0 | 1 | 6 | 0 | 1 | 6 | 1 | | MEMORY_MB | 1.0 | 1 | 7936 | 512 | 1 | 7936 | 256 | | DISK_GB | 1.0 | 1 | 28 | 0 | 1 | 28 | 1 | +----------------+------------------+----------+----------+----------+-----------+-------+------+ Recently this API was also added to Horizon (the web gui for Openstack) as hypervisor API and the compute db behind it is not preferred. Best wishes Lajos Katona (lajoskatona) Saeed <s.fazl73(a)gmail.com> ezt írta (időpont: 2024. máj. 10., P, 13:12): > Hello, > > Thanks a lot, but I didn't find the proper way for this. > > In CLI, I run *openstack hypervisor show hypervisor.domain.com > <http://hypervisor.domain.com>*, it then shows all the needed information > and data. > I run *openstack service list*, I see this: > > > +----------------------------------+------------------+-------------------------+ > | ID | Name | Type > | > > +----------------------------------+------------------+-------------------------+ > | 0a289a58588d4d00ac27ef92da8128de | keystone | identity > | > | 2cdf38eca0c0467fa856f1f1d91f8de3 | heat | orchestration > | > | 44393da92532478abd2dc1bdba1454ce | nova_legacy | compute_legacy > | > | 4cee55fd6c524134b010c65a7d406857 | ironic | baremetal > | > | 4f47c9bd69a34b5a9e0ef149823b7785 | neutron | network > | > | 59053bba36ed4944aaac7e8eb0f9ac01 | freezer | backup > | > | 670cb0faa6c64f8f82f0d19315e085a5 | trove | database > | > | 828ba6c2709349d597e469bf7fa55766 | glance | image > | > | 8c2671e5ddde47e4ae1accd2ce2eb696 | placement | placement > | > | 8c6357e12abb40e5a46b10be5f950c79 | heat-cfn | cloudformation > | > | c211ea65a3d440d1bb2add6f74423056 | watcher | infra-optim > | > | ceba54fa82f543579b245569b9fc127e | masakari | instance-ha > | > | d5c2952cc6904a0385cb20e5d39d1afe | nova | compute > | > | e238634b10db492a942e27714ee9d9b1 | gnocchi | metric > | > | f545381f4aa649a39122a258f48833fc | ironic-inspector | > baremetal-introspection | > > +----------------------------------+------------------+-------------------------+ > > If I do understand your reply correctly, you mean I should run a single > command for each server to find its free memory or CPU? For example if I > have 100 servers, I should run 100 times a command with different server > names? > > On Fri, 10 May 2024 at 10:33, Lajos Katona <katonalala(a)gmail.com> wrote: > >> Hi, >> Instead of Compute API, Placement is used for such statistics, see the >> API reference: >> https://docs.openstack.org/api-ref/placement/ >> >> You have to follow the resource providers on your compute, and resource >> providers can be in a tree with a root provider. >> For the available resources (resource classes) you have to list the >> resource provider inventories, and for >> the used resource you have to follow the allocations. >> Check the API ref and see the possibilities with CLI first perhaps. Be >> careful Placement API is microversioned like Compute API. >> >> From python SDK you can find the relevant methods here: >> >> https://opendev.org/openstack/openstacksdk/src/branch/master/openstack/plac… >> >> Best wishes >> Lajos Katona (lajoskatona) >> >> Saeed <s.fazl73(a)gmail.com> ezt írta (időpont: 2024. máj. 9., Cs, 19:06): >> >>> Hello, >>> >>> This is my python code: >>> >>>> import openstack >>> >>> openstack.enable_logging(debug=False) >>> >>> hypervisors = [] >>> >>> conn = openstack.connect(cloud='region1') >>> >>> for hypervisor in conn.compute.hypervisors(details=True): >>> >>> hypervisors.append(hypervisor.to_dict()) >>> >>> >>> I also run this: >>> >>>> for hypervisor in hypervisors: >>> >>> print(hypervisor) >>> >>> >>> This is the last result (other results are similar this): >>> >>>> {'cpu_info': None, 'host_ip': '1.1.1.1', 'hypervisor_type': 'QEMU', >>>> 'hypervisor_version': 6002000, 'name': 'some.fqdn.com', >>>> 'service_details': {'id': 'some-random-id', 'host': 'some.fqdn.com', >>>> 'disabled_reason': 'test'}, 'servers': None, 'state': 'up', 'status': >>>> 'disabled', 'uptime': ' 00:27:11 up 31 days, 10:14, 2 users, load >>>> average: 3.52, 3.20, 3.36\n', 'current_workload': None, 'disk_available': >>>> None, 'local_disk_used': None, 'local_disk_size': None, 'local_disk_free': >>>> None, 'memory_used': None, 'memory_size': None, 'memory_free': None, >>>> 'running_vms': None, 'vcpus_used': None, 'vcpus': None, 'id': >>>> 'some-random-id', 'location': Munch({'cloud': 'region1', 'region_name': >>>> 'Region1', 'zone': None, 'project': Munch({'id': 'somerandomid', 'name': >>>> None, 'domain_id': None, 'domain_name': None})})} >>> >>> >>> As you see, values like *cpu_info*, *disk_available*, *local_disk_used* >>> , *local_disk_size*, *local_disk_free*, *memory_used*, etc are *None*. >>> >>> How can I have the following values? In the Openstack UI I see it says >>> the free available disk, memory and CPU, but I don't know how to have them >>> in python. >>> >>> I see here >>> <https://docs.openstack.org/openstacksdk/latest/user/resources/compute/v2/hy…> >>> the above values, but for me all are None. >>> >>> Regards, >>> Saeed >>> >>

3 2

[manila][cinder][glance][nova][devstack][qa] Add ashrod98 to devstack-plugin-ceph core
by Goutham Pacha Ravi 10 May '24

10 May '24

Dear Stackers, I'd like to nominate Ashley Rodriguez to join the devstack-plugin-ceph core team. She has made significant contributions to Ceph integration and testing, especially with the Manila project. Ashley's work includes CI fixes and improvements to the cephadm installer, enabling multi-node installations and integrating ceph's new ingress service in our test jobs. She has also tackled tech debt in Manila's Ceph testing. If there are no objections, I'll add Ashley to the gerrit group next week. Ashley, your hard work is truly appreciated! Best regards, Goutham

4 4

[horizon] Next two weekly meetings cancellation
by Tatiana Ovchinnikova 10 May '24

10 May '24

Hello team, I'm on holiday for the next two weeks, so I'm cancelling weekly meetings on May 15 and 22. See you on May 29. Regards, Tatiana

1 0

[neutron] Drivers meeting canceled next Friday, May 17th
by Brian Haley 10 May '24

10 May '24

Hello Neutrinos: I'm going to preemptively cancel next Friday's drivers meeting as I will be traveling all week. I will be available intermittently on irc, so please send me an email if you need a quicker answer or if I don't respond to pings. -Brian

1 0