| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-07-04 | |||
| 10:39:54 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Use enum value instead of string service name https://review.openstack.org/476538 | |
| 10:40:18 | gibi | sfinucan: hi! I reworked https://review.openstack.org/#/c/476538/ according to your and Matt's comments | |
| 10:40:39 | openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient master: Fix cropping the endpoint url https://review.openstack.org/479818 | |
| 10:41:00 | sfinucan | gibi: I'm in the middle of a novaclient docs refactor for the docs-migration spec, but if you ping me in an hour or so, I'll take a look :) | |
| 10:41:45 | gibi | sfinucan: thanks, I will ping you later :) | |
| 10:45:42 | sfinucan | off the top of someone's head, is it OS_PROJECT_NAME or OS_TENANT_NAME that's deprecated? | |
| 10:54:02 | openstackgerrit | Stephen Finucane proposed openstack/python-novaclient master: doc: Create directory structure for docs migration https://review.openstack.org/480135 | |
| 10:54:02 | openstackgerrit | Stephen Finucane proposed openstack/python-novaclient master: doc: Switch from oslosphinx to openstackdocstheme https://review.openstack.org/480136 | |
| 10:55:03 | openstackgerrit | Stephen Finucane proposed openstack/python-novaclient master: tools: Remove dead script https://review.openstack.org/480138 | |
| 11:01:05 | openstackgerrit | Ildiko Vancsa proposed openstack/nova master: WIP/POC Implement new attach Cinder flow https://review.openstack.org/330285 | |
| 11:46:21 | openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient master: Fix cropping the endpoint url https://review.openstack.org/479818 | |
| 11:50:16 | openstackgerrit | OpenStack Proposal Bot proposed openstack/nova master: Imported Translations from Zanata https://review.openstack.org/477091 | |
| 11:50:48 | openstackgerrit | Béla Vancsics proposed openstack/nova master: Transform live_migration_post_dest notification https://review.openstack.org/469784 | |
| 11:57:20 | jangutter | sean-k-mooney: regarding https://review.openstack.org/#/c/460278, Netronome is scheduled to be upstreamed in kernel 4.13... It would be nice if we can get this consistent. | |
| 12:05:40 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Fix 'project-id' 'user-id' as required in server group https://review.openstack.org/480162 | |
| 12:18:06 | openstackgerrit | Alex Xu proposed openstack/nova master: placement: filter usage records by resource provider id https://review.openstack.org/480130 | |
| 12:18:48 | moshele | jangutter: in that case please push the required changes for netronme and I will check you patches on mellanox NIC | |
| 12:20:16 | jangutter | moshele: thanks, I'm just waiting for daylight to hit the guys in the USA. Sorry I was out of touch for so long. I claim infection and some convoluted timetables. | |
| 12:20:53 | moshele | jangutter: np | |
| 12:22:55 | jangutter | moshele: in a previous iteration of our legacy (non-TC) patches, we also had a step where we renamed the representor to the "tapXXXXX" name provided by Neutron. This helped OVSDB integration with OpenDaylight in our case. | |
| 12:26:40 | jangutter | moshele: In addition, it allowed us an easy way to persist across reboots and helped with migration. Would you guys consider such a step? | |
| 12:31:33 | moshele | jangutter: I don't understand why you need the rename, it SR-IOV so the migration is done with the pci device and then nova will look on a different hypervisor the represnetor which is bind to the new VF | |
| 12:32:09 | gibi | sfinucan: ping, if you have time please look at https://review.openstack.org/#/c/476538/ . thanks! | |
| 12:35:00 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Transform instance.resize_revert notification https://review.openstack.org/396811 | |
| 12:37:46 | sfinucan | gibi: Yup, I'll take a look now | |
| 12:37:51 | gibi | sfinucan: thanks | |
| 12:38:22 | openstackgerrit | Vladyslav Drok proposed openstack/nova master: Fix a typo in neutron network API docstring https://review.openstack.org/480167 | |
| 12:38:52 | jangutter | moshele: I might be remembering a very old version we had, where Neutron assumed that the vif_name would be the same on the remote host's bridge. | |
| 12:40:03 | jangutter | moshele: the agent listened for the plug event on OVSDB with the same vif_name and metadata. | |
| 12:45:25 | jangutter | moshele: but you're right, if the mechanism driver / agent doesn't look at things like openflow ports and vif_names, then the interface metadata in external-ids should allow it to link. | |
| 12:45:26 | moshele | jangutter: sorry I don't follow | |
| 12:46:02 | moshele | jangutter: see my comment in https://review.openstack.org/#/c/460278/ | |
| 12:46:39 | moshele | jangutter: can you follow the way In looking the representor port? | |
| 12:47:53 | jangutter | moshele: yes, I believe we just need 2 changes to work "out of the box", and neither should impact you. | |
| 12:48:52 | jangutter | moshele: we went through the code today, but the main developers are in USA time, so I just want to confirm that I'm not making up stuff. | |
| 12:50:03 | openstackgerrit | Stephen Finucane proposed openstack/os-vif master: doc: Remove cruft from releasenotes conf.py https://review.openstack.org/480092 | |
| 12:50:34 | moshele | jangutter: can you open a bug for https://github.com/openstack/os-vif/blob/master/vif_plug_ovs/linux_net.py#L272 I will fix it | |
| 12:51:46 | jangutter | moshele: so, the first change is that we also have a PF representor, and both are in physfn/net in sysfs. If you iterate through all netdevs there and find the first representor, it should fix that issue for us. | |
| 12:52:19 | jangutter | moshele: I'll do so. I just want to confirm on our side that our behaviour is intended. | |
| 12:53:53 | jangutter | moshele: the second issue is deriving the VF number from phys_port_name. I'll also verify our exact view on the stability of that interface. | |
| 12:58:35 | openstackgerrit | Ildiko Vancsa proposed openstack/nova master: WIP: Allow multi-attach in compute api https://review.openstack.org/271047 | |
| 12:58:37 | openstackgerrit | Ildiko Vancsa proposed openstack/nova master: libvirt: Allow multiple volume attachments https://review.openstack.org/267587 | |
| 12:59:25 | moshele | jangutter: I meant for the phys_port_name, the PF represnetor you can provide the code because I don't have a way to test it | |
| 13:06:22 | openstackgerrit | Alex Xu proposed openstack/nova master: placement: add traits support in the AllocationCandidates object https://review.openstack.org/479766 | |
| 13:06:23 | openstackgerrit | Alex Xu proposed openstack/nova master: placement: add traits support to get resource_providers with shared https://review.openstack.org/478464 | |
| 13:08:39 | ildikov | there's a quite frequent placement API functional test failure with the following error: nova.exception.TraitNotFound: No such trait MISC_SHARES_VIA_AGGREGATE. | |
| 13:09:01 | ildikov | is this something new or anyone is looking into this? | |
| 13:12:41 | alex_xu | ildikov: have you try after this patch https://review.openstack.org/#/c/478564/ | |
| 13:13:46 | ildikov | alex_xu: I rebased my chain I think yesterday, so I should have that one | |
| 13:14:01 | ildikov | alex_xu: it fails on the gate only, I haven't seen it locally | |
| 13:14:21 | ildikov | alex_xu: http://logs.openstack.org/85/330285/83/check/gate-nova-tox-functional-py35-ubuntu-xenial/5ea5cd0/console.html | |
| 13:16:18 | alex_xu | ildikov: emm...looks like still have something wrong | |
| 13:16:36 | ildikov | alex_xu: sorry for the bad news :( | |
| 13:31:48 | openstackgerrit | Stephen Finucane proposed openstack/nova master: trivial: Use valid UUIDs in test_admin_password https://review.openstack.org/393711 | |
| 13:32:39 | sfinucan | alex_xu, bauzas: Easy +2 here, if you fancy it https://review.openstack.org/#/c/393711/ (it's been there since Nov 2016) | |
| 13:33:04 | bauzas | oh yeah | |
| 13:34:10 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: trivial: Use valid UUIDs in test_admin_password https://review.openstack.org/393711 | |
| 13:34:53 | sean-k-mooney | jangutter: yes it would be good to enable it but it may have to wait till queens. 4.13 will not be out before then end of pike | |
| 13:35:28 | openstackgerrit | Stephen Finucane proposed openstack/nova master: rbd: Remove unnecessary 'encode' calls https://review.openstack.org/412356 | |
| 13:35:34 | bauzas | sfinucan: I stupidly +Wd the patch without waiting for a second +2 | |
| 13:35:49 | bauzas | sfinucan: hence me providing a new PS for not having the change to the gate pipr | |
| 13:35:57 | alex_xu | sfinucan: done | |
| 13:36:06 | sfinucan | bauzas: :) I have also done that more than once | |
| 13:36:14 | sfinucan | alex_xu, bauzas: Thanks folks :) | |
| 13:38:34 | openstackgerrit | Merged openstack/nova master: Fix 'project-id' 'user-id' as required in server group https://review.openstack.org/480162 | |
| 13:40:16 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-manage: Deprecate '--version' parameters https://review.openstack.org/453808 | |
| 13:40:16 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-manage: Rename 'api_db', 'cellsv2' commands https://review.openstack.org/453809 | |
| 13:44:56 | openstackgerrit | Alex Xu proposed openstack/nova master: placement: ensure MISC_SHARES_VIA_AGGREGATE trait created in the test https://review.openstack.org/480198 | |
| 13:45:44 | alex_xu | bauzas: sfinucan ^ still can see the 'MISC_SHARES_VIA_AGGREGATE' not found in the functional tests. I can't find the root cause, that patch is a choice which fallback to ensure the trait created first. | |
| 13:46:45 | bauzas | alex_xu: okay will look | |
| 13:46:59 | bauzas | alex_xu: I'm confused, we already have this trait ? | |
| 13:47:26 | alex_xu | bauzas: yea, it is in the os-traits lib, and it should be synced into the placement db | |
| 13:47:47 | bauzas | alex_xu: but computes wouldn't provide this trait now, right? | |
| 13:48:11 | sfinucan | alex_xu: This is the thing causing all the functional test failures, right? | |
| 13:48:12 | bauzas | I agree with that trait being in the *api* db given we now sync | |
| 13:48:33 | bauzas | but I don't remember when we merged to see computes pushing this trait | |
| 13:48:53 | bauzas | that said, I was (and I'm a bit still) out of my keyboard during last weeks | |
| 13:49:06 | alex_xu | sfinucan: yea, 29 hits http://logstash.openstack.org/#/dashboard/file/logstash.json?query=message:%5C%22%2Fusr%2Flocal%2Fbin%2Fdevstack-version:%20line%203:%20unexpected%20EOF%20while%20looking%20for%20matching%20%60''%5C%22%20AND%20tags:%5C%22console%5C%22&from=7d | |
| 13:49:24 | alex_xu | bauzas: the compute didn't push this trait | |
| 13:49:51 | alex_xu | bauzas: https://review.openstack.org/#/c/480198/1/nova/tests/functional/api/openstack/placement/fixtures.py@285 | |
| 13:50:06 | alex_xu | the fixture set that trait on the RP | |
| 13:50:51 | bauzas | alex_xu: sec | |
| 13:50:53 | alex_xu | bauzas: introduced by https://review.openstack.org/475448 | |
| 13:52:09 | bauzas | wait, I'm confused | |
| 13:52:22 | bauzas | what we provide as a functional test doesn't mean that Nova uses that | |
| 13:52:57 | jangutter | sean-k-mooney: acknowledged. Support should land at a similar time as Open vSwitch 2.8 stabilizes. | |
| 13:53:08 | bauzas | alex_xu: just to be clear, there are 2 things here | |
| 13:53:15 | bauzas | alex_xu: #1 is the list of traits | |
| 13:53:23 | bauzas | that's synced by os-traits, agreed | |
| 13:53:26 | bauzas | but | |
| 13:53:30 | bauzas | there is a #2 point | |
| 13:53:45 | bauzas | alex_xu: the traits mapping with ResourceProviders | |
| 13:54:01 | bauzas | alex_xu: and that, I don't think we merged it, correct? | |
| 13:54:25 | bauzas | ie. the fact that RP <uuid1> is mapped with trait <uuid2> | |
| 13:55:02 | sean-k-mooney | jangutter: will the netronome solution work without modification to ovs. my understnading was that ye did not support the TC flower protocol | |
| 13:55:08 | alex_xu | bauzas: for testing the shared resource provider, jay's patch fake a mapping https://review.openstack.org/#/c/475448/24/nova/tests/functional/api/openstack/placement/fixtures.py@283 | |
| 13:55:33 | bauzas | alex_xu: sure, but that's just for the functional test, right?N | |
| 13:55:42 | alex_xu | bauzas: yea, just for functional test | |
| 13:55:58 | bauzas | tbc, just for a *fixture*, even not a test | |
| 13:56:07 | bauzas | meaning that we don't even assert something | |