Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-11-11
15:30:32 gibi and I tried creating a neutron port with --no-fixed-ip
15:30:34 sean-k-mooney gibi: no
15:30:47 sean-k-mooney as far as i am aware we never enabled it
15:31:08 gibi yeah, I see Port 6370c969-73ca-41df-8c15-d91f10d44a59 requires a FixedIP in order to be used. (HTTP 400) with --no-fixed-ip ports
15:31:53 sean-k-mooney some of the concerns is that security groups and some qos policies may now work correctly
15:32:07 sean-k-mooney also the metadata api woudl not work
15:33:11 sean-k-mooney for l3 only backend like callico you also would have no network connectivty as you cant fall back to l2 in that case
15:33:40 gibi so I guess this cannot be added to nova generally, we have to somehow special case it if we need it
15:34:00 sean-k-mooney we now have the connectivity in the port
15:34:12 sean-k-mooney it will be either legacy l2 or l3 i belive
15:34:24 sean-k-mooney so we could allow it for l2 ports and maybe legacy
15:34:32 sean-k-mooney legacy was for backend that did not say
15:34:39 sean-k-mooney but your milage would vary
15:34:52 gibi thanks
15:35:18 sean-k-mooney actully rodolfos patch already has that https://review.opendev.org/#/c/669411/1/nova/network/neutronv2/api.py
15:36:33 gibi so that handles the l3 only backend case, what can we do with secgroups?
15:37:04 sean-k-mooney am well secgroups almost always deals with l3+
15:37:20 sean-k-mooney so if you have no ip then you can only use the ethertype rule
15:37:32 sean-k-mooney neutron will still do mac spoofing protection
15:37:41 sean-k-mooney i dont think we need to care about that
15:37:56 sean-k-mooney well we can care we just dont need to do anything
15:38:16 gibi OK, so we could repropos rodolfo's patch if we want
15:38:27 sean-k-mooney yep i have no issue with doing that
15:38:35 sean-k-mooney ie proceedign with it
15:38:46 sean-k-mooney he already did the neutron feature
15:38:53 sean-k-mooney which is what it was waiting on really
15:39:02 sean-k-mooney https://review.opendev.org/#/c/645645/
15:39:52 sean-k-mooney now that ovn is in tree we might want to make sure that reporst the attibute as l2 but other then that i dont see a blocker here
15:41:09 sean-k-mooney i think its kind of expected that if you have an adressless port that you cant use things like floating ips which neutron should block for us
15:41:11 gibi sean-k-mooney: thanks. I might repropose that at some point as it seem there is downstream need for it
15:41:48 sean-k-mooney we had a downstream request too at one point but it was obviously not a high priorit since its still pending
15:42:46 sean-k-mooney ovn is also updated fyi https://github.com/openstack/neutron/blob/master/neutron/plugins/ml2/drivers/ovn/mech_driver/mech_driver.py#L171
15:43:02 sean-k-mooney so it looks like everything is in place on the neutron side
15:43:28 gibi coolio
15:43:43 gibi I still not get the priority info hence the conditional from my side too
15:43:49 gibi let see
15:44:14 zigo sean-k-mooney: It IS included, it's just done in openstack-pkg-tools automatically.
15:45:08 zigo Also, nova-compute.conf only contains a few hypervisor specifics for switching virtualization depending on which nova-compute-FOO you install.
15:45:33 zigo ie: nova-compute-qemu will install qemu as virt driver, nova-compute-kvm will do kvm, etc.
15:45:40 zigo But everything else is in nova.conf ...
15:52:31 sean-k-mooney zigo: ya i saw they were minimal
15:52:51 sean-k-mooney so ideally you would do one of two things move the db and api db config into there own files
15:53:25 sean-k-mooney or move all compute node relevent options into the compute one
15:55:31 sean-k-mooney zigo: as they currently are you cannot use the auto rpc version since rocky
15:56:24 sean-k-mooney i.e. this https://github.com/openstack/nova/blob/master/nova/conf/upgrade_levels.py#L56-L57
18:48:22 openstackgerrit Ghanshyam Mann proposed openstack/nova master: DNM: Testing system scope in tempest https://review.opendev.org/740124
19:50:44 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Migrate nova-grenade-multinode job to zuulv3 native https://review.opendev.org/742056
21:21:33 openstackgerrit Merged openstack/nova stable/victoria: Fix virsh domifstat to get vhostuser vif statistics https://review.opendev.org/760684
21:21:44 openstackgerrit Merged openstack/nova master: Improve error handling during service level check https://review.opendev.org/762175
21:52:32 openstackgerrit Daniel de Oliveira Pereira proposed openstack/nova master: Avoid getting transport_url from config file when updating Cell0 https://review.opendev.org/762433
22:16:04 openstackgerrit Merged openstack/nova master: Update contributor guide for Wallaby https://review.opendev.org/754427
23:26:59 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Migrate nova-grenade-multinode job to zuulv3 native https://review.opendev.org/742056
23:34:20 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Add functional-py39 testing https://review.opendev.org/760884
23:36:01 gmann stephenfin: ^^ need to re-approve this. I rebased it to resolve merge conflict.
#openstack-nova - 2020-11-12
01:20:58 openstackgerrit chengsheng proposed openstack/nova master: Add hypervisor CPU feature check during live migration https://review.opendev.org/762330
02:32:20 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Migrate nova-grenade-multinode job to zuulv3 native https://review.opendev.org/742056
02:45:55 openstackgerrit chengsheng proposed openstack/nova master: Modify the default value of the force parameter in live migration https://review.opendev.org/762458
05:28:19 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Migrate nova-grenade-multinode job to zuulv3 native https://review.opendev.org/742056
05:57:57 openstackgerrit Xinran WANG proposed openstack/nova-specs master: SRIOV Enabled Nic Support Specification https://review.opendev.org/742785
06:14:33 openstackgerrit norman shen proposed openstack/nova-specs master: Adds spec for instance live resize https://review.opendev.org/141219
06:50:56 openstackgerrit Jorhson Deng proposed openstack/nova master: To deal instance with soft-deleting in _init_instance https://review.opendev.org/761264
07:54:21 openstackgerrit Jorhson Deng proposed openstack/nova master: To deal instance with soft-deleting in _init_instance https://review.opendev.org/761264
08:07:18 f0o Hi, not sure if this is an issue in nova/kvm or the guest-os: Sometimes (really random) when a linux guest, predominantly CoreOS/Flatcar, attempts to reboot the instance is placed into Puased state instead. However the pause state is erroneously because there's no resume action possible so something must've gone bonkers in the state sync. Only issueing Hard-Reboot via cli will actually
08:07:20 f0o bring the guest back alive
08:23:17 openstackgerrit Balazs Gibizer proposed openstack/nova stable/victoria: Improve error handling during service level check https://review.opendev.org/762471
08:45:00 bauzas good and very cold morning Nova
09:19:22 lyarwood has anyone written up a bug for functional failing (due to an eventlet bug) on 3.9 yet?
09:19:25 lyarwood AttributeError: 'Semaphore' object has no attribute '_at_fork_reinit'
09:19:51 lyarwood https://github.com/eventlet/eventlet/issues/646 fixed it a few days ago
09:26:35 lyarwood kay weird now it's working
09:26:39 lyarwood okay*
09:26:55 lyarwood nvm
09:29:21 openstackgerrit Wenping Song proposed openstack/nova master: Remove redundant check https://review.opendev.org/762482
09:48:38 openstackgerrit chengsheng proposed openstack/nova master: Modify the default value of the force parameter in live migration https://review.opendev.org/762458
09:53:54 stephenfin gibi: sean-k-mooney1: FYI https://review.opendev.org/762396
09:54:04 stephenfin that's the OSC change to support shelve offload
10:05:53 gibi stephenfin: ack, thanks, will look
10:12:30 openstackgerrit Brin Zhang proposed openstack/nova master: Cyborg shelve/unshelve support https://review.opendev.org/729563
10:31:51 openstackgerrit Jorhson Deng proposed openstack/nova master: To deal instance with soft-deleting in _init_instance https://review.opendev.org/761264
10:33:36 openstackgerrit Balazs Gibizer proposed openstack/nova master: Revert "Temporarily disable parts of heal port allocation test" https://review.opendev.org/755995
10:55:09 openstackgerrit Brin Zhang proposed openstack/nova master: [Trivial] Rename host to hostname in conductor manager https://review.opendev.org/762499
10:56:31 brinzhang_ gibi: I would like to rename the variable 'host' express 'hostname' in conductor manager, is it ok?
10:59:31 gibi brinzhang_: just quickly looked at it it seems OK, you might consider renaming node to nodename in the same context
11:02:09 brinzhang_ gibi: ok, I will do it later, thanks
11:06:35 gibi fyi folks, there is a new type of gate failure (at least it is new for me) https://bugs.launchpad.net/nova/+bug/1903979
11:06:35 openstack Launchpad bug 1903979 in OpenStack Compute (nova) "nova-live-migration job fails during evacuate negative test" [High,Confirmed]
11:12:16 lyarwood gibi: I can look at this today
11:12:45 gibi lyarwood: thanks, I'm pushing a e-r signature for it
11:14:26 gibi here is the e-r patch https://review.opendev.org/762502
11:16:03 lyarwood urgh zuul's webui isn't responding for me
11:16:09 gibi it is slow
11:16:12 gibi for me
11:16:16 gibi but eventually opens
11:16:33 lyarwood gibi: do you have the raw console log link?
11:19:04 gibi https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_ca4/762319/1/check/nova-live-migration/ca4ccb9/job-output.txt
11:20:28 gibi lyarwood: ^^
11:23:19 sean-k-mooney1 stephenfin: it more or less looks ok but i left a few comments
11:23:39 sean-k-mooney1 stephenfin: i think you are failing out to early
11:23:49 openstackgerrit Brin Zhang proposed openstack/nova master: [Trivial] Rename host/node to hostname/nodename in conductor manager https://review.opendev.org/762499
11:35:28 stephenfin sean-k-mooney1: replied

Earlier   Later