Friday, 2026-06-26

gmaanstephenfin: melwitt nova change for enforce_scope testing also ready https://review.opendev.org/c/openstack/nova/+/98694601:59
opendevreviewKamil Sambor proposed openstack/nova master: Add concurrency_backend config option for Nova services  https://review.opendev.org/c/openstack/nova/+/99376010:44
chandank`sean-k-mooney: Hello, Are you still reviewing the nvme spec? Reason I am asking because I need to update the cleanup state diagram and few nit pics. Or I will wait for your feedback.11:42
chandank`sorry wrong channel11:49
*** chandank` is now known as chandankumar11:50
opendevreviewUdayendu Kar proposed openstack/python-novaclient master: tests: skip osprofiler test when dependency is not installed  https://review.opendev.org/c/openstack/python-novaclient/+/99504812:42
opendevreviewmitya-eremeev-2 proposed openstack/nova stable/2025.1: Replace eventlet.event.Event with threading.Event  https://review.opendev.org/c/openstack/nova/+/99506313:40
stephenfingibi: sean-k-mooney: Could one of you sanity check this for me? https://review.opendev.org/c/openstack/python-openstackclient/+/995003/13:41
stephenfinI assume it's fine but I've never done anything dual-stack'y like that13:41
sean-k-mooneyyou can have both if you have subnets for both on the same network13:43
sean-k-mooneybut ill take a look now13:43
sean-k-mooneyi am not sure if we can spcify both in nova api but if you pass a newuton port or network the the port can be assigne both13:44
sean-k-mooneystephenfin: hum so accordign to the api ref  13:45
gibiI will rely on sean-k-mooney's eyes as I also never did dual stack13:45
sean-k-mooneynetworks.fixed_ip 13:45
sean-k-mooneyis ipv4 only13:45
sean-k-mooneyA fixed IPv4 address for the NIC. Valid with a neutron or nova-networks network.13:45
sean-k-mooneyhttps://docs.openstack.org/api-ref/compute/#id1113:46
sean-k-mooneyi will have to check the code but lookign at the fixed_ip filed there i dont see evnicne we have ever supprot ipv6 in server create13:47
sean-k-mooneyi.e  where we provide an ipv6 adress to use13:47
sean-k-mooneyit will owrk if you pass port or netork that has an ipv6 subnet13:47
sean-k-mooneybut we dont currently allow you to specify a subnet to create the prot on or pass an ipv6 ip13:48
sean-k-mooneyso either the api ref is wrong (*that is possibel) or ipv6 is not suprpot in genral so the issue is not about dualstack or not at that point13:49
sean-k-mooneystephenfin: ok so looking at the openapi schmea you wote13:51
sean-k-mooneyteh api accpate networks.fixed_ip but we allow that to be an ipv4 or v6 adress13:52
sean-k-mooneyat least in the schema13:52
sean-k-mooneybut we only support one of fixed_ip prot or uuid per network13:53
sean-k-mooneymeainign you cant pass both for the same port13:53
sean-k-mooneyhttps://github.com/openstack/nova/blob/master/nova/api/openstack/compute/schemas/servers.py#L283-L29913:53
stephenfinI wonder can we test it somehow just to see what happens?13:54
stephenfinI imagine deploying dual-stack devstack is not a trivial matter 😅13:54
sean-k-mooneyi dont think its that terirbel you just need to create a newton networ with both13:55
sean-k-mooneythe host does not need it13:55
sean-k-mooneybut i can see if i have a devstack i can test it on quickly13:55
sean-k-mooneyi think the failure mode is we dont accpet muleipel fixed ip pre netowrk object in our api. we could be i think that is why that is there13:59
sean-k-mooneyso it snot someabout v4 or v613:59
fricklerI tested it locally and indeed the v6 address does not get set13:59
sean-k-mooneyits more that we only supprot 1 adress per port that we create14:00
fricklerwhile without a fixed-ip specified, I do get both addr types14:00
sean-k-mooneyfrickler: ya it woudl change the subnet we attach too14:00
sean-k-mooneyi think before we make the client change we woudl need a new api microverion but in the past we said if you neded somethign like ethat you shoudl precreate the neutorn port14:01
sean-k-mooneybasiclly because we didnt wnat to keep ading proxy apis for that14:01
sean-k-mooneyfrickler: i assume you created the vm with curl?14:01
sean-k-mooneyor did you just use the propsoed patch?14:02
fricklerI did use the patch. against a 2024.1 admittedly14:03
frickler*cloud14:03
sean-k-mooneydevstack default to a dual stack deployment https://paste.opendev.org/show/bq6cSQIBviTCgXyGkvA7/ so il test it again on that deployment which si masteris with the patch14:04
sean-k-mooneyfrickler: so what im seeitn tiwth the patch applied is regardless of the oder only the v4 adress gets applied14:20
sean-k-mooneyhttps://paste.opendev.org/show/bzPsu7xQOJ0oc0fzJzSs/14:20
sean-k-mooneyalthough the vms get both v4 and v6 adresses  the v6 adress is just not the one i specified14:21
sean-k-mooneylooking at the reques boday this is what is being sent https://paste.opendev.org/show/baVOCdYizNOYiccHJmWe/14:24
sean-k-mooneystephenfin: so the patch does not actlly work14:24
sean-k-mooneyit remvoes the check but the client is still only seingi the ipv4 adress becuase i think we check the ipv4 first in the code and then dont check the ipv6 adress14:25
sean-k-mooneystephenfin: i marked https://bugs.launchpad.net/python-openstackclient/+bug/2152032 as invlid feel free to change that if you think that is wrong14:33
sean-k-mooneyfrickler: thanks for checking :)14:34
stephenfinno, that's great. thanks for looking into it :) frickler++ sean-k-mooney++14:37
opendevreviewSylvain Desgrais proposed openstack/nova master: Allow security group names with leading/trailing spaces  https://review.opendev.org/c/openstack/nova/+/99507314:37
fricklerwell ... if nova accepts both v4 and v6 in the API, but then ignores one of those, arguably that's a bug that could/should get fixed an not an API change?14:37
fricklerusing a preconfigured port works, but has more drawbacks, too. like security groups also need to get applied manually to the port and are ignored during the server create call14:38
opendevreviewSylvain Desgrais proposed openstack/nova master: Allow security group names with leading/trailing spaces  https://review.opendev.org/c/openstack/nova/+/99507314:39
opendevreviewBalazs Gibizer proposed openstack/nova master: Limit the port fields during security_group query  https://review.opendev.org/c/openstack/nova/+/99489614:49
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Detect AMD SEV-SNP support  https://review.opendev.org/c/openstack/nova/+/99476414:57
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Support launchSecurity element for SEV-SNP  https://review.opendev.org/c/openstack/nova/+/99492014:57
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Omit optional fields for SEV/SEV-ES  https://review.opendev.org/c/openstack/nova/+/99492114:57
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Drop sev capability check in launch instance flow  https://review.opendev.org/c/openstack/nova/+/99492914:57
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Launch instances with SEV-SNP memory encryption  https://review.opendev.org/c/openstack/nova/+/99493014:57
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Drop sev capability check in launch instance flow  https://review.opendev.org/c/openstack/nova/+/99492915:21
opendevreviewTakashi Kajinami proposed openstack/nova master: AMD SEV: omit iommu='on' for virtio devices  https://review.opendev.org/c/openstack/nova/+/90963515:21
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Launch instances with SEV-SNP memory encryption  https://review.opendev.org/c/openstack/nova/+/99493015:21
opendevreviewTakashi Kajinami proposed openstack/nova master: Update documentations for AMD SEV-SNP support  https://review.opendev.org/c/openstack/nova/+/99509015:51
opendevreviewTakashi Kajinami proposed openstack/nova master: Update documentations for AMD SEV-SNP support  https://review.opendev.org/c/openstack/nova/+/99509016:13
opendevreviewMerged openstack/nova master: Add regression test for bug 2062127  https://review.opendev.org/c/openstack/nova/+/97665417:22
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Launch instances with SEV-SNP memory encryption  https://review.opendev.org/c/openstack/nova/+/99493017:44
opendevreviewTakashi Kajinami proposed openstack/nova master: Update documentations for AMD SEV-SNP support  https://review.opendev.org/c/openstack/nova/+/99509017:44
opendevreviewTakashi Kajinami proposed openstack/nova master: Remove [libvirt] num_memory_encrypted_guests  https://review.opendev.org/c/openstack/nova/+/99512018:02
opendevreviewTakashi Kajinami proposed openstack/nova master: Remove [libvirt] num_memory_encrypted_guests  https://review.opendev.org/c/openstack/nova/+/99512018:02
opendevreviewAdam Harwell proposed openstack/placement master: Add ASGI application entry point for Placement API  https://review.opendev.org/c/openstack/placement/+/99512919:06
opendevreviewAdam Harwell proposed openstack/placement master: Add ASGI application entry point for Placement API  https://review.opendev.org/c/openstack/placement/+/99513319:10
rm_work[m]erk19:11
sean-k-mooneyfrickler: osc is not sending both. nova will accpet an ip in v4 or v6 format but it wont accpet multiple of them19:54
sean-k-mooneyfrickler: i have not tested it with curl but i woudl expect a 400 band request if you tried to repoeate the file of make it a list instaed of a stirng19:54
sean-k-mooneyin the json payload19:54
melwittgmaan: question on https://review.opendev.org/c/openstack/nova/+/986946 let me know if I am wrong :)20:02
melwittsean-k-mooney: why did you +2 this? :) it's phrased as if it's an addition to someone else's patch but I don't see other related patches? oh, it's that you requested testing in the patch above and added it below https://review.opendev.org/c/openstack/nova/+/99443720:17
gmaanmelwitt: checking, I am not sure why that test does not exist in base test class and only in scope testing20:19
melwittyeah, it was different than everything else seems like20:19
opendevreviewAdam Harwell proposed openstack/placement master: Add ASGI application entry point for Placement API  https://review.opendev.org/c/openstack/placement/+/99513320:21
gmaanmelwitt: nice catch, replied the reason. it was added when policy was SYSTEM_READER and we did not move it to base class when removed the system scope from our policy20:27
gmaanI will fix it20:27
melwittah gotcha. cool I will look again after you update20:27
opendevreviewGhanshyam Maan proposed openstack/nova master: Update policy scope enforcement testing  https://review.opendev.org/c/openstack/nova/+/98694620:38
opendevreviewSylvain Desgrais proposed openstack/nova master: Allow security group names with leading/trailing spaces  https://review.opendev.org/c/openstack/nova/+/99507320:40
opendevreviewTakashi Kajinami proposed openstack/nova master: Remove [libvirt] num_memory_encrypted_guests  https://review.opendev.org/c/openstack/nova/+/99512021:43
opendevreviewMerged openstack/nova master: Add regression tests for interface attach MTU bug  https://review.opendev.org/c/openstack/nova/+/99443722:29
opendevreviewMerged openstack/nova master: Fix network mtu metadata disappears after attaching  https://review.opendev.org/c/openstack/nova/+/93609322:35

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!