| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-04-03 | |||
| 08:03:37 | tetsuro | Without this change, how do NRPs that don’t have requested resources appear in prov_sum? Are we going to expose them without resource information? | |
| 08:04:12 | tetsuro | Could you tell me why we need all the NRPs in the tree? If there aren’t specific and concrete usecase so far, I even don’t think we have to support that all NRP in prov_sum since the design depends on the usecase. | |
| 08:04:27 | tetsuro | I’m also asking the same question in https://review.openstack.org/#/c/556873/. | |
| 08:10:25 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: [placement] api-ref: Fix parameters https://review.openstack.org/558142 | |
| 08:29:54 | bauzas | good morning Novaers | |
| 08:35:30 | fanzhang | oh hi bauzas, sorry to bother. I accidentally found this trivial bug, https://review.openstack.org/#/c/558414/, should I add CI manually to reviewer or just wait? Kind a new to post patch to nova. :) | |
| 08:36:25 | bauzas | fanzhang: in general, we try to avoid small nit fixes like this one | |
| 08:37:28 | fanzhang | bauzas yeah, I do think it's not appropriate. | |
| 08:37:44 | bauzas | fanzhang: to answer your original question, that change is in the queue http://zuul.openstack.org/ | |
| 08:38:04 | bauzas | I can see it there in the check pipeline | |
| 08:38:29 | bauzas | so you just need to wait until all the jobs are done so that you'd get a Zuul review | |
| 08:39:43 | bauzas | fanzhang: you can have some docs in https://docs.openstack.org/infra/manual/developers.html#automated-testing | |
| 08:40:47 | fanzhang | bauzas oh, cool. Thanks. I see there are lots of CI jobs in nova. But we only have a few in trove. I thought maybe nova is not the same as trove. :) | |
| 08:41:09 | kashyap | alex_xu_: Thanks for reviewing this! https://review.openstack.org/#/c/534384/ | |
| 08:42:28 | fanzhang | bauzas as for the nit fix above, should I abandon it or ? | |
| 08:47:05 | fanzhang | I noticed there was one trivial fix like above, see https://review.openstack.org/#/c/558183/ . So anything I should do you can just help point it out, thanks :) I do want to know how nova team try to avoid small nit fixes. | |
| 08:51:31 | johnthetubaguy | fanzhang: we have a list of "trival bug fixes" on this etherpad: https://etherpad.openstack.org/p/rocky-nova-priorities-tracking | |
| 08:58:51 | fanzhang | johnthetubaguy thanks, looks great :) | |
| 08:59:28 | kashyap | Is this the magic incantation to "recheck" Citrix XenServer CI? -- 'xenserver: recheck' | |
| 09:01:19 | kashyap | johnthetubaguy: Meanwhile; this is ready, also has a +2: https://review.openstack.org/#/c/534384/22 | |
| 09:01:43 | openstackgerrit | jichenjc proposed openstack/nova master: z/VM Driver: Spawn and destroy function of z/VM driver https://review.openstack.org/527658 | |
| 09:08:10 | openstackgerrit | Merged openstack/nova master: Fix a missing white space in exception message https://review.openstack.org/558183 | |
| 09:10:49 | openstackgerrit | Merged openstack/os-vif stable/pike: ovs: do not delete port if already exists https://review.openstack.org/550080 | |
| 09:17:11 | johnthetubaguy | kashyap: ah, good call out, will take another peak at that ASAP | |
| 09:41:02 | kashyap | Splendid! | |
| 09:43:21 | kashyap | johnthetubaguy: Oh, darn, so it is the spacing that is causing the output to be all mangled? -- https://review.openstack.org/#/c/534384/22/nova/conf/libvirt.py@546 | |
| 09:43:24 | kashyap | Good catch! | |
| 09:45:19 | kashyap | johnthetubaguy: Any suggestion to fix that? As it stands, indentation doesn't render | |
| 09:45:22 | kashyap | stephenfin: You about? | |
| 09:45:31 | stephenfin | kashyap: I am | |
| 09:45:44 | kashyap | stephenfin: When you are: I wonder if you have any tricks to get this right: https://review.openstack.org/#/c/534384/22/nova/conf/libvirt.py@546 | |
| 09:46:00 | kashyap | See the rendering, the indentation is gone, and the config is highlighted: http://logs.openstack.org/84/534384/22/check/build-openstack-sphinx-docs/016a796/html/configuration/config.html | |
| 09:46:05 | johnthetubaguy | kashyap: some of it used to, got standup right now, so distracted | |
| 09:46:25 | kashyap | stephenfin: `grep` for the the string: "This allows specifying granular CPU feature flags when specifying" to arrive at the right place in the HTML | |
| 09:46:28 | johnthetubaguy | kashyap: hmm, looks like what I had working is totally broken again | |
| 09:46:41 | stephenfin | kashyap: It's a bug in oslo_config.sphinxext https://review.openstack.org/#/q/status:open+project:openstack/oslo.config+branch:master+topic:bug/1755783 | |
| 09:46:55 | johnthetubaguy | "* ..." and "\n" used to render OK, but its a bit broken right now | |
| 09:47:13 | kashyap | stephenfin: Ah, interesting | |
| 09:47:45 | stephenfin | kashyap: ...which should be fixed soon as those patches are merged. This is the one you care about https://review.openstack.org/553860 | |
| 09:47:58 | kashyap | stephenfin: So I'll add a pointer to your patch series (thanks!). I don't need to make any changes yet, do I? | |
| 09:48:11 | stephenfin | kashyap: Once that is merged though, there are a couple of things that still need changing. I'll comment now | |
| 09:49:55 | kashyap | stephenfin: Ah, thank you. I'll subscribe to that 'oslo_config' patch series | |
| 09:51:42 | kashyap | stephenfin: Hmm, hope that'd be stable branches backports compatible; because I'll be backporting that change to two branches | |
| 09:52:01 | kashyap | (As it's addressing an important guest perf regression) | |
| 09:52:08 | stephenfin | kashyap: You're just going to lose some of the docs formatting there, I'm afraid | |
| 09:52:18 | stephenfin | It'll still render but poorly | |
| 09:52:40 | kashyap | stephenfin: For backports, you mean? | |
| 09:52:59 | stephenfin | Indeed | |
| 09:53:18 | kashyap | Maybe I can adjust it according to the equivalent syntax for backports? (Assuming there is such.) | |
| 09:54:01 | stephenfin | kashyap: Not sure there's any way to do it but I guess it can be experimented with, aye | |
| 09:54:02 | stephenfin | o/ | |
| 10:18:59 | johnthetubaguy | kashyap: why did we decide to exclude host-model in that patch again? I get excluding host-passthrough. | |
| 10:21:55 | sean-k-mooney[m] | o/ | |
| 10:29:25 | sean-k-mooney[m] | "spare time" before proposing them upstream. | |
| 10:29:25 | sean-k-mooney[m] | hi wierd question but would the nova team object to me runnning a personal ci against nova that is triggred by a specific comment. i nolonger have direct control over what is run in the intel nfv ci and there are a few things i would like to test perodically that cannot currently be tested upstream. similarly there are a few things that i think can be tested upstream that i want prototype with zuul v3 in my | |
| 10:30:05 | sean-k-mooney[m] | melwitt: stephenfin dansmith jaypipes ^ | |
| 10:30:22 | stephenfin | sean-k-mooney[m]: If it's not voting, I don't see why not | |
| 10:31:31 | sean-k-mooney[m] | stephenfin: ya it would not set the verifed flag to +/-1 just comment back sucess/failure when triggered by a comment | |
| 10:32:22 | sean-k-mooney[m] | i was also going to restict it to people with intel emails/the nova core team. | |
| 10:33:47 | sean-k-mooney[m] | i am setting up some ci infra for internal use and wantted to tag this on the side to do things like test livemigration with cpu pinning/hugepgaes/vhost-user since that is currently disabled in the intel nfv-ci | |
| 10:34:16 | sean-k-mooney[m] | im going to follow up with them to trun that back on but you know its nice to be able to check some of those edgecases in an automated way | |
| 10:34:53 | stephenfin | Indeed it is. That sounds like a sane thing to do from my perspective anyway | |
| 10:38:49 | sean-k-mooney[m] | this is currently an inovation time/personal time thing so i might not have that much time to work on it but i set up nodepool over the weekend and should have zuul runing by the end of the week. with zuul v3 its not that hard to trigger build on both internal stuff and external. i was thinking of trying to add a patch work trigger to zuul too at somepoint but ya just taught i would ask. | |
| 10:53:48 | openstackgerrit | Merged openstack/nova master: [Trivial]Add missing blank space in conf description https://review.openstack.org/558414 | |
| 10:58:37 | openstackgerrit | Merged openstack/nova master: Fix unit tests to work with new oslo.config https://review.openstack.org/558084 | |
| 11:11:08 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: Update the cells FAQs and scheduler maintenance docs. https://review.openstack.org/556459 | |
| 11:11:08 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: Add --enable and --disable options to nova-manage update_cell https://review.openstack.org/555416 | |
| 11:54:25 | openstackgerrit | Michael Still proposed openstack/nova master: Move create_tap_dev to the new utility method. https://review.openstack.org/557835 | |
| 11:54:26 | openstackgerrit | Michael Still proposed openstack/nova master: Move set_vf_interface_vlan to the new utility module. https://review.openstack.org/558487 | |
| 12:03:03 | jaypipes | sean-k-mooney[m]: I have no issues with that at all. I think it would be great. | |
| 12:12:44 | bauzas | sahid: soft -1 for https://review.openstack.org/#/c/511188/ for asking you to be clear about all the possible values for the new option that you're going to add | |
| 12:13:02 | bauzas | sahid: once you're done, ping me and jaypipes and then I'll review it quickly | |
| 12:20:50 | alex_xu_ | jaypipes: dansmith fried_bunny edleafe bauzas, good morning, Sundar has a summary for the fpga based on our previous discussion at https://etherpad.openstack.org/p/Cyborg-Nova-Multifunction, he is looking for some review and feedback | |
| 12:21:11 | johnthetubaguy | stephenfin: would be great if you could double check my thinking on this runway patch: https://review.openstack.org/#/c/552242 | |
| 12:22:37 | openstackgerrit | Chris Dent proposed openstack/nova master: [placement] Parse forbidden traits in query strings https://review.openstack.org/556819 | |
| 12:22:37 | openstackgerrit | Chris Dent proposed openstack/nova master: [placement] Filter allocation candidates by forbidden traits in db https://review.openstack.org/556660 | |
| 12:22:38 | openstackgerrit | Chris Dent proposed openstack/nova master: [placement] Support forbidden traits in API https://review.openstack.org/556820 | |
| 12:24:43 | openstackgerrit | sean mooney proposed openstack/nova master: add mtu to libvirt xml for ethernet and bridge types https://review.openstack.org/553072 | |
| 12:25:52 | sean-k-mooney[m] | sahid: stephenfin added teh extra unit tests to ^ and rebased it against master. otherwise its the same as the previous version. | |
| 12:34:20 | fried_bunny | alex_xu_: Roger that; it's been on my list. | |
| 12:34:41 | alex_xu_ | fried_bunny: thanks | |
| 12:35:35 | efried | The pagan rites of spring are over. | |
| 12:44:38 | kashyap | johnthetubaguy: Hi there; was AFK. Will answer on the review, but to answer your question here: | |
| 12:45:10 | kashyap | johnthetubaguy: For 'host-model' + PCID doesn't make sense in this case — If QEMU already supports PCID, it would be enabled by 'host-model'. And if it is not supported, adding it doesn't make it magically appear :-) | |
| 12:45:16 | kashyap | johnthetubaguy: Does the above answer you? | |
| 12:49:23 | kashyap | johnthetubaguy: We logged a warning, instead of an _exception_ due to the prior discussion on the review: https://review.openstack.org/#/c/534384/20/nova/virt/libvirt/driver.py | |
| 12:56:08 | bhagyashris | johnthetubaguy, alex_xu_, melwitt: Hi, addressed review comments could you please review the same https://review.openstack.org/#/c/511825/ thank you :) | |
| 12:58:37 | kashyap | stephenfin: Hey there, so ... some OCD discussion on captialization: | |
| 12:58:59 | kashyap | stephenfin: On your comment here: https://review.openstack.org/#/c/534384/22/nova/conf/libvirt.py@568 | |
| 12:59:29 | alex_xu_ | bhagyashris: johnthetubaguy melwitt, good to have agreement on this first http://lists.openstack.org/pipermail/openstack-dev/2018-March/128914.html | |
| 12:59:46 | johnthetubaguy | kashyap: I got the impression host model had PCID missing put the CPU might still support it? | |
| 12:59:49 | kashyap | stephenfin: My rule to capitalize the sentence following a colon: if there is just _one_ sentence after colon, use lower case; otherwise upper case. | |
| 13:00:19 | johnthetubaguy | kashyap: maybe we should add a TODO to raise once its not a backport? | |
| 13:00:21 | kashyap | stephenfin: Hope that make sense? | |
| 13:00:24 | kashyap | johnthetubaguy: Hi | |
| 13:01:00 | stephenfin | kashyap: Fine by me. That should have been preceded by 'nit'. Does seem it's purely stylistic too (which I didn't know) https://www.grammarly.com/blog/capitalization-after-colons/ | |
| 13:01:11 | kashyap | johnthetubaguy: No-no: With `host-model`, *if* you have the updated microcode, guest/host kernel, QEMU and libvirt, then libvirt will do the Right Thing. | |
| 13:01:34 | kashyap | stephenfin: No worries, I enjoy the English nit-picks. :-) | |
| 13:02:35 | johnthetubaguy | kashyap: ah, good. I got the wrong end of the stick about that bit then. | |
| 13:02:36 | kashyap | johnthetubaguy: After the backports are done -- we will remove the restriction on 'host-model' | |