| opendevreview | Doug Goldstein proposed openstack/nova master: Add trunk subports to network metadata https://review.opendev.org/c/openstack/nova/+/941227 | 00:55 |
|---|---|---|
| *** mhen_ is now known as mhen | 03:02 | |
| *** ykarel_ is now known as ykarel | 05:50 | |
| opendevreview | Noppakorn Kaewsalabnil proposed openstack/python-novaclient master: Adapt for api 2.89 in volume resource https://review.opendev.org/c/openstack/python-novaclient/+/937642 | 06:58 |
| *** mdfr5 is now known as mdfr | 07:13 | |
| opendevreview | Thibaut Démaret proposed openstack/nova master: libvirt: add disk rotation_rate support for local disks https://review.opendev.org/c/openstack/nova/+/979693 | 09:19 |
| gibi | bauzas: https://review.opendev.org/q/topic:%22bug/2168874%22 | 10:02 |
| bauzas | gibi: noted, that will be my meat for this morning | 10:06 |
| gibi | bauzas: also one more context from yesterday that good to know https://review.opendev.org/c/openstack/nova/+/1008841/2#message-225a4587c643274bf73fcb9517e853a9f1800321 we discussed with Dan that there is a slight inefficiency in the late affinity server group check during live migration that can only be fixed with an RPC bump to send request_spec to compute during live migration. We now think that it | 10:23 |
| gibi | make sense to do that now in CCLM work for the new RPC calls as that anyhow need a bump | 10:23 |
| bauzas | ah ok | 10:27 |
| bauzas | I could help if needed | 10:27 |
| gibi | I think Dan takes this into account in the his RPC patch | 10:31 |
| gibi | so this is only context for that review | 10:31 |
| bauzas | k | 10:32 |
| opendevreview | Balazs Gibizer proposed openstack/nova master: Func reproducer for bug/2168874 https://review.opendev.org/c/openstack/nova/+/1007917 | 10:52 |
| opendevreview | Balazs Gibizer proposed openstack/nova master: Fix cross-cell cold migration with PCI devices https://review.opendev.org/c/openstack/nova/+/1008071 | 10:52 |
| opendevreview | Balazs Gibizer proposed openstack/nova master: Reload source instance before cross-cell cloning https://review.opendev.org/c/openstack/nova/+/1015204 | 10:52 |
| opendevreview | Balazs Gibizer proposed openstack/nova master: Reload source instance before cross-cell cloning https://review.opendev.org/c/openstack/nova/+/1015204 | 12:53 |
| stephenfin | gibi: didn't check IRC yesterday but I don't have anything to add after https://review.opendev.org/c/openstack/nova-specs/+/1008081/comments/d44b4866_83621ea1 | 13:47 |
| gibi | stephenfin: OK thanks | 13:54 |
| opendevreview | Merged openstack/nova stable/2026.2: tests: Functional reproducer for bug 2088831 https://review.opendev.org/c/openstack/nova/+/1013898 | 14:23 |
| opendevreview | Takashi Kajinami proposed openstack/nova master: Replace constant by constants from errno module https://review.opendev.org/c/openstack/nova/+/1015358 | 14:41 |
| opendevreview | Takashi Kajinami proposed openstack/nova master: libvirt: Verify mulitpath check during init_host https://review.opendev.org/c/openstack/nova/+/1015362 | 14:52 |
| tkajinam | sean-k-mooney, ^^^ this is probably a light-weight approach using unit tests | 14:53 |
| sean-k-mooney | ya more or less | 14:54 |
| sean-k-mooney | what i was actuly expecting you to assert was that _check_multipath | 14:54 |
| sean-k-mooney | is called wehn you call init_hopst | 14:54 |
| sean-k-mooney | since that is actully waht we want to assert rather then the side effect | 14:55 |
| sean-k-mooney | so iw as expect one new test that mocked _cehck_multipath and asserted it was caleld in respocne too drvr.init_host('dummyhost') | 14:55 |
| sean-k-mooney | is_multipath_running is lookg through the implation of _check_multipath into the internals | 14:56 |
| sean-k-mooney | which si not really how unit test should be wtiien for the init_host funciton | 14:56 |
| tkajinam | I can add that, though that felt a bit redundant to me. | 14:56 |
| sean-k-mooney | well your curertn test are asserting based on os-brick internslas | 14:57 |
| sean-k-mooney | so thee are sort of fragile | 14:57 |
| sean-k-mooney | but ya that was the minimal check that i was expectign since that was the inventain twe wanted to maintian | 14:57 |
| tkajinam | no these tests are not touching its internal really | 14:58 |
| tkajinam | the tests mocks the exact method nova calls so it's not going into the logic within os-brick | 14:58 |
| sean-k-mooney | they are mocking the is multipath runnign property on the os-brick class | 14:58 |
| sean-k-mooney | that is very much the internals | 14:58 |
| tkajinam | ok | 14:59 |
| sean-k-mooney | how they are curerntly moecked can be valed for testing drvr._check_multipath() | 14:59 |
| sean-k-mooney | but its not really corect for testing init_host | 15:00 |
| sean-k-mooney | so i woudl revert the change you are makeign | 15:00 |
| sean-k-mooney | and just add the new test that mocks _check_multipath() | 15:00 |
| sean-k-mooney | and assert its called | 15:00 |
| sean-k-mooney | and then we shoudl be good | 15:00 |
| sean-k-mooney | init_host tests shoudl nto know about os-brick sicne that is hidden behidn _check_multipath if that makes sense | 15:01 |
| tkajinam | ok | 15:01 |
| tkajinam | ah we already have test_init_host_checks_ip | 15:01 |
| tkajinam | doing a similar thing ... | 15:01 |
| tkajinam | ok then I can follow that ... | 15:01 |
| sean-k-mooney | oh i made the mistake of clickign who all lines on that file... | 15:03 |
| opendevreview | Takashi Kajinami proposed openstack/nova master: libvirt: Verify mulitpath check during init_host https://review.opendev.org/c/openstack/nova/+/1015362 | 15:05 |
| tkajinam | sean-k-mooney, updated ^^^^ | 15:06 |
| sean-k-mooney | yep +2 | 15:06 |
| tkajinam | it seems unit tests started showing multiple deprecation warnings from pyroute2 ... | 15:07 |
| tkajinam | hm it seems we just have to get a new version of pyroute2 https://github.com/svinota/pyroute2/commit/6940f07e52dc15b99d78bde4ed6bff392d14db88 | 15:10 |
| tkajinam | though I remember it's pinned due to nova/os-vif problem | 15:11 |
| opendevreview | Alberto Estévez Caldas proposed openstack/nova-specs master: libvirt: Add SEV-SNP HOST_DATA from user data https://review.opendev.org/c/openstack/nova-specs/+/1015367 | 15:12 |
| opendevreview | Lucas HENRY proposed openstack/nova master: libvirt: Stop the event threads in Host.cleanup() https://review.opendev.org/c/openstack/nova/+/1010011 | 15:57 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!