September 2025 - openstack-discuss

[kolla][kolla-ansible] Python dependency handling in custom Ansible module without toolbox
by sarah.kh09＠gmail.com 22 Sep '25

22 Sep '25

Hi everyone, I'm developing a new Ansible module for one of the kolla-ansible roles, and it depends on a Python package that needs to be installed via pip on the target nodes. I understand that Kolla Ansible generally avoids installing Python packages directly on hosts, and prefers handling such dependencies through the kolla_toolbox container. However, in my case: - Deploying a sidecar container isn't feasible within our current architecture - Delegating execution to localhost or kolla_toolbox would introduce complexity that we’d rather avoid Given these considerations, I'm looking for alternative ways to handle this dependency while staying aligned with Kolla's principles of reproducibility and container isolation. Has anyone faced a similar situation or found a clean workaround? Is using kolla-toolbox the only standard and recommended approach in such cases? Thanks a lot for your help and feedback!

2 2

[nova] Nova API call duration increases after upgrading to Caracal
by Chang Xue 21 Sep '25

21 Sep '25

Hello, we recently upgraded from Bobcat to Caracal. After upgrading, an interesting increment of duration of nova api call is observed. It's involved to multiple API calls, e.g. os-query-sets, /v2.1/servers/detail, etc. And it seems to be accumulated, meaning whenever nova-api is restarted, the duration seems to be fairly fast, giving response within like 80, 90 ms. As it keeps running for like up to 10 days, the duration increases to over 500 ms. I noticed that Caracal is using UnifiedLimitsDriver instead of DbQuotaDriver. I tried to change nova.conf to use DbQuotaDriver but it seems not working well with reducing the duration. I understand that calling to Keystone and Placement (and likely other external APIs) could result to longer duration, but pretty confused why it shows up as accumulated. Wondering if anyone else in Openstack community noticed the similar issue, thanks!

6 14

[aodh][ceilometer] how create alarms with prometheus ceilometer backend
by Nguyễn Hữu Khôi 21 Sep '25

21 Sep '25

Hello. I try to set up alarm such as: openstack alarm create --type prometheus --type prometheus --name memory_high_alarm --query 'memory_usage{resource_id="21d0792e-2d01-4df9-958a-d9018d13207f"}' --threshold 10 I use ceilometer with prometheus backend but when it shows: +---------------------------+------------------------------------------------------------------+ | Field | Value | +---------------------------+------------------------------------------------------------------+ | alarm_id | cb7b8d07-0af1-4771-92c8-1a4c2ce4d1d1 | | name | memory_high_alarm | | description | prometheus alarm rule | | enabled | True | | ok_actions | [] | | alarm_actions | [] | | insufficient_data_actions | [] | | repeat_actions | False | | type | prometheus | | time_constraints | [] | | project_id | 31ef81adc8ad406e9e91e3eb8776c959 | | user_id | 2e7a252719434d93ae0d270c8753294d | | timestamp | 2025-09-18T14:10:50.467568 | | state | *insufficient data * | | state_timestamp | 2025-09-18T14:10:50.467568 | | state_reason | Not evaluated yet | | severity | low | | evaluate_timestamp | 2025-09-18T14:10:50.485384 | | comparison_operator | eq | | threshold | 10.0 | | query | memory_usage{resource_id="21d0792e-2d01-4df9-958a-d9018d13207f"} | +---------------------------+------------------------------------------------------------------+ and in aodh logs: 2025-09-18 21:20:02.442 1079 INFO aodh.evaluator [-] initiating evaluation cycle on 1 alarms 2025-09-18 21:20:02.442 1079 WARNING aodh.evaluator [-] Skipping alarm cb7b8d07-0af1-4771-92c8-1a4c2ce4d1d1, unsupported type: prometheus I am using openstack 2025.1 I would like to get some guides on this, Many thanks. Nguyen Huu Khoi

2 6

[aodh]how the evaluation process works
by Nguyễn Hữu Khôi 21 Sep '25

21 Sep '25

Hello, I would like to understand how the evaluation process works. For example, I have this rule: openstack alarm create --name memory_use_test \ --type gnocchi_resources_threshold \ --metric memory.usage \ --threshold 102 \ --comparison-operator gt \ --aggregation-method mean \ --granularity 60 \ --evaluation-periods 3 \ --resource-id dca36039-c603-422c-8366-f6ba4e9539c3 \ --resource-type instance Does this mean that when the alarm-evaluator service is running per a period, it will check the last 3 data points, each aggregated over 60 seconds? Thank you. Regards. Nguyen Huu Khoi

2 2

[ptl][charms][tc] Need a PTL volunteer for OpenStack Charms project team
by Goutham Pacha Ravi 20 Sep '25

20 Sep '25

Hello Stackers, We're missing a Project Team Lead (PTL) for the OpenStack Charms project team for the "2026.1" (Gazpacho) release cycle. If you're interested, please let us know by proposing a PTL appointment (e.g. [1]). Thanks, Goutham Pacha Ravi (gouthamr) [1] https://review.opendev.org/c/openstack/governance/+/943254

2 2

[release] Release countdown for week R-1, Sep 22 - 26
by Elõd Illés 19 Sep '25

19 Sep '25

Development Focus ----------------- We are on the final mile of the 2025.2 "Flamingo" development cycle! Remember that the 2025.2 "Flamingo" final release will include the latest release candidate (for cycle-with-rc deliverables) or the latest intermediary release (for cycle-with-intermediary deliverables) available. September 26th, 2025 is the deadline for final 2025.2 "Flamingo" release candidates as well as any last cycle-with-intermediary deliverables. We will then enter a quiet period until we tag the final release on October 1st, 2025. Teams should be prioritizing fixing release-critical bugs, before that deadline. Otherwise it's time to start planning the 2026.1 "Gazpacho" development cycle, including discussing Forum and PTG sessions content, in preparation of OpenInfra Summit Europe (Paris-Saclay, France; October 17th-19th, 2025) and virtual OpenStack PTG (October 27th-31st, 2025). Actions ------- Watch for any translation patches coming through on the stable/2025.2 branch and merge them quickly. If you discover a release-critical issue, please make sure to fix it on the master branch first, then backport the bugfix to the stable/2025.2 branch before triggering a new release. Please drop by #openstack-release with any questions or concerns about the upcoming release ! Upcoming Deadlines & Dates -------------------------- Final 2025.2 "Flamingo" release: October 1st, 2025 OpenInfra Summit Europe: October 17th-19th, 2025 OpenStack PTG (virtual): October 27th-31st, 2025 Előd Illés irc: elodilles @ #openstack-release

1 0

[neutron] VPN service/router stays PENDING_CREATE - 2024.2 Dalmatian
by Shubham Yadav 19 Sep '25

19 Sep '25

Hello OpenStack Community, I'm working with OpenStack VPNaaS (using strongSwan as the backend driver) and have a question about tunnel configuration capabilities. Current Setup: OpenStack VPNaaS with strongSwan driver (StrongSwanDriver) Single VPN Service (gateway) with multiple IPSec Site Connections Each connection targets different remote peers with distinct subnets Configuration Details: VPN Service: Connected to a single router Multiple IPSec Site Connections under the same VPN Service: Connection 1: To MY office (10.10.100.0/24) via peer 68.x.y.34 Connection 2: To Rackspace (172.24.60.0/22) via peer 184.n.p.53 Shared IKE and IPSec policies across connections Local endpoint groups covering multiple subnets (192.168.0.0/24, 10.43.0.0/16) Question: Is it supported/recommended to have multiple IPSec Site Connections under a single VPN Service? I understand this creates multiple tunnels from the same OpenStack router to different remote peers. Current Issue: One of my connections shows "PENDING_CREATE" status while the other is "ACTIVE". I'm trying to determine if this is: A configuration issue on my end A limitation of having multiple tunnels per VPN service A strongSwan-specific behavior in the OpenStack context Specific Questions: Does VPNaaS officially support multiple IPSec connections per VPN Service? Are there any known limitations or best practices when using multiple tunnels? Should each remote site have its own dedicated VPN Service instead? I've checked the documentation but couldn't find clear guidance on multi-tunnel scenarios. Any insights from the community would be greatly appreciated. Environment: OpenStack version: 2024.2 Dalmatian Neutron VPNaaS with strongSwan backend Kubernetes-deployed OpenStack (OpenStack-Helm) Waiting for your reply, Thank you for your time and assistance! Best regards, Shubham

1 0

[all][elections][ptl] 2026.1 Gazpacho: PTL Election Conclusion & Results
by Sławek Kapłoński 18 Sep '25

18 Sep '25

Hi, Thank you to the electorate, to all those who voted and to all candidates who put their name forward for Project Team Lead (PTL) in this election. A healthy, open process breeds trust in our decision making capability. Thank you to all those who make this process possible. Now for the results of the PTL election process, please join me in extending congratulations to the following PTLs: * Adjutant - Dale Smith (dalees) * Barbican - Mauricio Harley (mharley) * Blazar - Pierre Riteau (priteau) * Cinder - Jon Bernard (jbernard) * Cloudkitty - Rafael Weingartner () * Cyborg - Li Liu () * Designate - Omer Schwartz () * Glance - Cyril Roelandt () * Heat - Takashi Kajinami (tkajinam) * Horizon - Tatiana Ovchinnikova (tmazur) * Keystone - Artem Goncharov (gtema) * Kolla - Michal Nasiadka (mnasiadka) * Magnum - Dale Smith (dalees) * Manila - Carlos Silva (carloss) * Masakari - sam sue () * Mistral - Arnau Morin (amorin) * Neutron - Brian Haley (haleyb) * Nova - Rene Ribaud (Uggla) * Octavia - Michael Johnson (johnsom) * OpenStack Helm - Vladimir Kozhukalov (kozhukalov) * OpenStackAnsible - Dmitriy Rabotyagov (noonedeadpunk) * OpenStackSDK - Artem Goncharov (gtema) * Puppet OpenStack - Takashi Kajinami (tkajinam) * Quality Assurance - Ghanshyam Maan (gmaan) * Rally - Andrey Kurilin (andreykurilin) * Skyline - Wenxiang Wu (wu_wenxiang) * Storlets - Takashi Kajinami (tkajinam) * Sunbeam - Guillaume Boutry (gboutry) * Swift - Tim Burke (timburke) * Tacker - Yasufumi Ogawa (yasufum) * Telemetry - Matthias Runge (mrunge) * Trove - wu chunyang (wuchunyang) * Zaqar - Hao Wang (wanghao) * Zun - Hongbin Lu () Elections: * Horizon: https://civs1.civs.us/cgi-bin/results.pl?id=E_2cbcfa05fd0ec211 For remaining projects (without PTL candidates), TC will work on the appointment. Election process details and results are also available here: https://governance.openstack.org/election/ Thank you to all involved in the PTL election process, Slawek Kaplonski, on behalf of the OpenStack Technical Election Officials

2 1

PTG October 2025 Team List Announcement
by Kendall Nelson 18 Sep '25

18 Sep '25

Hello Everyone! The October 2025 Project Teams List is official! Projects + Teams - OpenInfra Project Teams - Kata Containers - StarlingX - - OpenStack - Services - Blazar - Cinder - CloudKitty - Ironic - Keystone - Kolla - Magnum - Manila - Neutron - Nova - Octavia - OpenStack Horizon - OpenStack-Ansible - Sunbeam - Swift - Telemetry - Other OpenStack Teams - Eventlet Removal - openstack-i18n - Release Management - Technical Committee If your team was planning to meet and isn’t in this list, please contact ptg(a)openinfra.dev IMMEDIATELY. Soon I will be contacting moderators to sign up for time via the PTGBot[1] once it is configured to accept those reservations. As usual, feel free to let us know if you have any questions! Thanks! -Kendall (diablo_rojo) [1] PTGBot Docs: https://opendev.org/openstack/ptgbot#open-infrastructure-ptg-bot [2] PTG Registration: http://ptg.openinfra.org/ <http://ptg.openinfra.o/g>

1 0

[heat][mistral][tc] Status of ply
by Takashi Kajinami 18 Sep '25

18 Sep '25

Hi, As you might know, some of the OpenStack projects, heat and mistral, use yaql to support flexible user input. The yaql library has been tiny and stable, thanks to the ply[1] library which is the python-written implementation of lex and yacc . However, it was declared that no further release of ply is planned[2] and the author now suggests copying it into the project[3]. The old released version has been working so far, but I'm unsure if it could survive through recent drastic changes in python package tooling. Looking at the code, it seems all codes are tagged with BSD 3-clause license[3] though the repository lacks the LICENSE file. I guess that the easiest approach is to import the codes to somewhere within yaql (like yaql._ply) and use the imported codes, with a note explaining these codes are imported from ply. This follows the suggestion by the author, but I'd like to get some guidance from TC to make sure it won't break the whole license expectation. An alternative approach is to implement our own parser, though I'm not too sure if that's feasible at least from my PoV. If we want this approach, then we need actual volunteer for it who will immediately start working on it. If neither of the two options above work, we might end up removing yaql functionality completely from OpenStack. AFAIK the one in heat was added for TripleO which was already retired so the impact may be limited, while for mistral removal could have lager impact. Thank you, Takashi [1] https://github.com/dabeaz/ply [2] https://github.com/dabeaz/ply?tab=readme-ov-file#important-notice---october… [3] https://github.com/dabeaz/ply?tab=readme-ov-file#questions-and-answers [4] https://github.com/dabeaz/ply/blob/master/src/ply/lex.py#L10-L33 -- Takashi Kajinami irc: tkajinam github: https://github.com/kajinamit launchpad: https://launchpad.net/~kajinamit

4 5