Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2022-02-24
09:05:06 yuval today is the deadline right?
09:05:21 gibi yuval: yes, today is the deadline
09:05:28 yuval ok
09:05:37 gibi but I think you are in a good position
09:06:12 yuval thank you
10:20:01 bauzas yuval: gibi: looking at it
11:00:55 bauzas sean-k-mooney: ha
11:01:12 bauzas you already +Wd the lightbits driver
11:01:20 bauzas I had a comment
11:01:35 bauzas we need a reno file
11:02:06 sean-k-mooney[m] ah you are right
11:02:15 sean-k-mooney[m] i can pull +w or we can add a followup patch
11:02:16 bauzas yuval: https://review.opendev.org/c/openstack/nova/+/821606/16/nova/tests/unit/virt/libvirt/volume/test_lightos.py#28
11:02:51 bauzas sean-k-mooney: well, given it's already going, and given the deadline is today, that's OK
11:03:04 bauzas sean-k-mooney: but I want yuval to create it
11:03:10 bauzas and also I had a comment https://review.opendev.org/c/openstack/nova/+/821606/16/nova/tests/unit/virt/libvirt/volume/test_lightos.py#28
11:03:24 bauzas yuval: around ?
11:03:40 sean-k-mooney[m] ok yuval if you add a follow up patch with a release note we can review that quickly
11:04:01 sean-k-mooney[m] ya just saw the question re flags
11:04:16 sean-k-mooney[m] i thihk that was a leftover form a previous patch
11:05:41 sean-k-mooney[m] oh its for the retires
11:05:53 sean-k-mooney[m] that is what they ment to set in the test
11:07:07 sean-k-mooney[m] although they asserted its default value in the test so not entirly needed
11:17:37 yuval hey
11:17:40 yuval I am here
11:17:57 yuval release note ok
11:18:08 bauzas thanks
11:18:17 bauzas yuval: do you know how to do it ?
11:18:58 yuval last time I add it in cinder I copied an existing release note and edited it.
11:19:11 yuval need to keep the correct format etc
11:19:34 sean-k-mooney[m] we have a tool
11:19:51 sean-k-mooney[m] you can do tox -e venv reno new lightos
11:20:05 sean-k-mooney[m] that will create a release note you can edit
11:20:22 sean-k-mooney[m] then you can test that its valied with tox -e releasenotes
11:20:43 bauzas yuval: no, let me give you the doc
11:21:18 yuval ok
11:21:28 sean-k-mooney[m] in this case you can delete all sections except feature where you should provide a short overview of the feature and any version requirements
11:21:40 sean-k-mooney[m] e.g. this need lightos version X
11:22:27 bauzas yuval: https://docs.openstack.org/nova/latest/contributor/releasenotes.html
11:22:58 bauzas yuval: just use the 'features' section for the YAML file
11:23:12 bauzas yuval: also, I had a nit for a test
11:23:23 bauzas yuval: could you just reply on gerrit ?
11:23:32 yuval yes, I saw I didnt understand it actually
11:23:43 bauzas https://review.opendev.org/c/openstack/nova/+/821606/16/nova/tests/unit/virt/libvirt/volume/test_lightos.py#28
11:24:31 sean-k-mooney[m] ok going to grab coffee. brb
11:31:28 gibi bauzas: I'm +2 on the unified limits series. Is it OK with you if I go back and plug the missing +As to the series?
11:31:45 bauzas gibi: sure
11:31:50 gibi ack, on it
11:31:51 bauzas do this if you want
11:53:00 sean-k-mooney :) i care about that landign but also know i do not have the time before FF to load the context required to review it properly so i have been staying away from the series
11:56:42 gibi melwitt: I have a small question in https://review.opendev.org/c/openstack/nova/+/789963
12:23:22 opendevreview Jonathan Race proposed openstack/nova master: driver/secheduler/docs for Adds Pick guest CPU architecture based on host arch in libvirt driver support https://review.opendev.org/c/openstack/nova/+/822053
12:23:23 opendevreview Jonathan Race proposed openstack/nova master: zuul-job for Adds Pick guest CPU architecture based on host arch in libvirt driver support https://review.opendev.org/c/openstack/nova/+/828372
12:24:36 chateaulav gibi: fixed the pep8 error, i promise this is the last time the driver patch will be updated, focus is on the zuul patch today
12:26:28 gibi chateaulav: no worries :)
12:32:49 sean-k-mooney given gibi has +2'd the first two patches i take it this series is generally in a good postion for review
12:33:01 gibi sean-k-mooney: yes
12:33:15 sean-k-mooney cool i will try an take a look at them shortly so
12:33:26 gibi the last patch adds the gate testing, that still has some failing tempest test but most of the tempest already green
12:33:38 gibi which is a good sign that the emulation actually working
12:35:00 sean-k-mooney ack do we want to hold +w until that last ci patch is green im not going to +w until we are happy with all 3 fo the feature patches as i think this should all merge at once anyway but is the zuul job part of the DOD
12:35:28 sean-k-mooney oh the feature is now only 2 patches
12:35:44 sean-k-mooney oh right the third is the ci patch
12:35:57 chateaulav sean-k-mooney: correct
12:36:30 chateaulav i figured once that one is green ill remove it from check, that way you guys can implement it the weekly and periodic\
12:37:19 sean-k-mooney honestly im ok with havign one emulated env in check
12:37:28 chateaulav ok
12:37:44 sean-k-mooney but if we want to test others i would put them in the periodic weekly pipeline
12:38:10 gibi sean-k-mooney: I'm OK to land the feature today even without the CI path, I think the current state of the CI result on the CI patch already looks good enough
12:38:12 sean-k-mooney its very unlikely that a change woudl break mips emulation but not arm
12:38:54 sean-k-mooney gibi: ack ok ill focus on the first too so then take a look at the ci patch
12:39:57 sean-k-mooney chateaulav: the m1.tiny flavor size might break some senario tests due to sapce if ti does i think m1.micro is big enough for arm cirros but small then m1.tiny
12:40:36 sean-k-mooney for now lets leave it as you have it and see if it breaks anything
12:40:37 chateaulav ok
12:42:23 yuval File "/home/yuval/workspace/opendev/nova/nova/virt/driver.py", line 129, in <module>
12:42:23 yuval "supports_remote_managed_ports": os_traits.COMPUTE_REMOTE_MANAGED_PORTS,
12:42:23 yuval AttributeError: module 'os_traits' has no attribute 'COMPUTE_REMOTE_MANAGED_PORTS'
12:42:31 yuval I get this while running unittests
12:42:40 yuval what I am doing wrong?
12:43:09 opendevreview Merged openstack/nova master: Lightbits LightOS driver https://review.opendev.org/c/openstack/nova/+/821606
12:43:10 gibi yuval: you need to rebuild you tox env
12:43:22 gibi yuval: it probably has an old os-traits lib
12:43:30 yuval tox -e ?
12:43:35 gibi yuval: try tox -r -e <env>
12:43:37 yuval or remove it
12:43:39 gibi the -r is the rebuild
12:43:43 yuval ok
12:43:44 gibi or you can remove .tox dir
12:43:46 yuval thanks
12:43:55 yuval yes I will try it if it wont work
12:44:06 sean-k-mooney yuval: ya this is likely just an old os-traits version
12:44:16 sean-k-mooney the other way to fix it is to manually install it
12:44:42 sean-k-mooney .tox/<env>/bin/python3 -m pip install -U os-traits
12:45:19 sean-k-mooney that is also how you can locally test with os-triats changes that are not released
12:45:51 gibi bauzas, gmann: I'v started re-reading the policy series and I have concerns about assisted volume snapshot https://review.opendev.org/c/openstack/nova/+/828994/7/nova/api/openstack/compute/assisted_volume_snapshots.py#42
12:45:52 sean-k-mooney just replace os-triats with the path to the git repo on disk
13:00:45 opendevreview Merged openstack/nova master: libvirt: Register defaults for undefined hw image properties https://review.opendev.org/c/openstack/nova/+/800708
13:00:54 opendevreview Merged openstack/nova master: manage: Add image_property commands https://review.opendev.org/c/openstack/nova/+/824392
13:02:55 yuval hey added release note
13:05:44 yuval I got a little mixed up with git issues

Earlier   Later