| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-05-30 | |||
| 14:15:38 | bauzas | I just provided a patch for that purpose because I thought it was nicier from an interface PoV | |
| 14:15:45 | jaypipes | bauzas: Sylvain, you were on the hangout where we discussed doing the instance_uuids parameter. | |
| 14:15:53 | jaypipes | bauzas: it was right before you went on PTO | |
| 14:16:02 | jaypipes | mriedem: yes | |
| 14:16:02 | bauzas | jaypipes: we agreed on making the claims | |
| 14:16:18 | bauzas | I don't remember if we agreed on the interface | |
| 14:16:51 | bauzas | anyway, looks like the general direction was on another trend | |
| 14:16:57 | bauzas | edleafe: I'll provide a PS9 | |
| 14:17:08 | bauzas | because I still have concerns with PS7 | |
| 14:17:17 | jaypipes | bauzas: yes, I'm pretty sure we agreed to pass the instance_uuids as a new param on RPC. I remember that because I had said changing the RequestSpec object definition was cleaner but was persuaded that wasn't a good idea to add non-DB fields to an object definition. | |
| 14:17:41 | mriedem | maybe we should make a rule that anything discussed while one of the 5 people coding this is out, that we actually communicate decisions back into the mailing list, or something :) | |
| 14:17:58 | jaypipes | mriedem: hindsight 20/20 | |
| 14:18:17 | bauzas | so | |
| 14:19:10 | bauzas | I'm not particularly tied to my PS8 | |
| 14:19:29 | bauzas | I just think https://review.openstack.org/#/c/469037/2 is worth it anyway | |
| 14:19:40 | mriedem | ok, good, so 1 reqspec and list of instance uuids over the rpc method, not within the reqspec | |
| 14:20:26 | mriedem | bauzas: isn't that https://blueprints.launchpad.net/nova/+spec/request-spec-use-by-compute which hasn't been started yet? | |
| 14:20:38 | bauzas | I'm just surprised we had a consensus on an interface that made it very tied with how we use placement | |
| 14:20:45 | bauzas | but okay | |
| 14:21:18 | bauzas | mriedem: yup, it can be supported by that bp | |
| 14:22:09 | mriedem | andymccr: circling back, ok so i see at least one tempest test failed - but it's unclear why really, so if there are tempest console logs in a separate file it'd be cool to see those to debug | |
| 14:22:15 | mriedem | andymccr: if you're working on that, ack | |
| 14:23:09 | andymccr | mriedem: got it. thanks! I'm running a test manually and i'll get some logs and circle back if i cant figure it out, but tl;dr the hostmappingnotfound error is not the cause of the issue | |
| 14:23:28 | jaypipes | bauzas: how does this interface tie us to placement? | |
| 14:23:31 | mriedem | right i didn't think it would be, just ugliness in the logs | |
| 14:24:02 | bauzas | jaypipes: because we expose as a parameter some attribute only needed because we want to use placement | |
| 14:24:27 | bauzas | jaypipes: the original interface is ask_me_for(this_request) | |
| 14:24:54 | bauzas | jaypipes: now it becomes ask_me_for_(this_request, those UUIDs) | |
| 14:25:09 | bauzas | but anyway, I don't want to open wounds | |
| 14:25:25 | mriedem | since we're not pulling nova-scheduler out, at this point i don't really care | |
| 14:25:32 | mriedem | nova-scheduler is internal to nova | |
| 14:25:48 | bauzas | mriedem: which is what wasn't said in PS7's commit msg... | |
| 14:26:24 | bauzas | if we're saying we don't care of the beauty of the scheduler API, I'm totally fine | |
| 14:26:42 | bauzas | then, we shouldn't be afraid of having alternatives returned by the scheduler... | |
| 14:27:07 | cdent | bauzas: we decided that we have to return alternatives. the only open question is on the data structure | |
| 14:27:46 | mriedem | btw sheeple, the bottom big ass counting quotas change has a +2 on it https://review.openstack.org/#/c/446239/ | |
| 14:27:50 | bauzas | cdent: the reason why I asked edleafe to abandon https://review.openstack.org/#/c/465767/5 in favor of https://review.openstack.org/#/c/465175/7 | |
| 14:28:22 | jaypipes | mriedem: I have the quotas one up in gertty. | |
| 14:28:32 | jaypipes | mriedem: since I previously reviewed it, I'll take that on. | |
| 14:44:09 | EmilienM | mriedem: hey, do you know how nova-coresec launchpad works? I created one for tripleo and don't know the next step | |
| 14:44:19 | EmilienM | should I add the group to another group? | |
| 14:44:53 | mriedem | EmilienM: i never set that up, but https://security.openstack.org/vmt-process.html might have details | |
| 14:44:58 | mriedem | otherwise i'd ask fungi | |
| 14:45:05 | EmilienM | ok, thx | |
| 14:45:11 | EmilienM | you have a checkbox "This bug is a security vulnerability" | |
| 14:45:12 | openstackgerrit | Chris Friesen proposed openstack/nova master: show flavor info in server details https://review.openstack.org/434514 | |
| 14:45:43 | cfriesen | api-ref tweaks ^ | |
| 14:45:46 | openstackgerrit | Jay Pipes proposed openstack/nova master: pci: add uuid field to PciDevice object https://review.openstack.org/469147 | |
| 14:57:47 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: Have scheduler claim resources https://review.openstack.org/463993 | |
| 14:57:47 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: WIP: Scheduler returning alternatives per destination https://review.openstack.org/465175 | |
| 14:57:48 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: Pass a list of instance UUIDs to scheduler https://review.openstack.org/465171 | |
| 15:01:03 | bauzas | jaypipes: mriedem: edleafe: there it is, restored the original direction for https://review.openstack.org/465171 with a bit of fixes from me | |
| 15:01:15 | openstackgerrit | Merged openstack/nova master: cinder: add attachment_create method https://review.openstack.org/459111 | |
| 15:03:36 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Completely remove mox from unit/network/test_linux_net.py https://review.openstack.org/445504 | |
| 15:04:33 | openstackgerrit | Merged openstack/nova master: cinder: add attachment_update method https://review.openstack.org/459113 | |
| 15:10:14 | sfinucan | jaypipes, mriedem: Could I ask you folks to take a look at https://review.openstack.org/#/q/status:open+project:openstack/nova+topic:bp/reserve-numa-with-pci today? | |
| 15:11:06 | jaypipes | sfinucan: sure thing. wrapping up a review on melwitt's quotas patch then will do. | |
| 15:11:21 | sfinucan | jaypipes: Sweet. Thank you :) | |
| 15:11:36 | jaypipes | sfinucan: in the meantime, I offer you https://review.openstack.org/469147\ | |
| 15:12:09 | sfinucan | didn't I review that already 🤔 | |
| 15:13:01 | sfinucan | jaypipes: Ah, I did - https://review.openstack.org/#/c/414768/. Yup, I'll take a look now | |
| 15:14:52 | jaypipes | sfinucan: ya, I broke that patch out of the n-r-p series because it's self-contained. | |
| 15:15:49 | bauzas | mriedem: I do wonder about a specific case for https://bugs.launchpad.net/nova/+bug/1689692 | |
| 15:15:49 | openstack | Launchpad bug 1689692 in OpenStack Compute (nova) "MarkerNotFound when limit>num(instances) and marker starts in cell0" [High,In progress] - Assigned to Matt Riedemann (mriedem) | |
| 15:16:10 | bauzas | mriedem: given 3 instances with #1 ACTIVE, #2 ERROR and #3 ACTIVE | |
| 15:16:43 | bauzas | mriedem: if we set the marker to #1, we're still able to get #2 and #3 AFAICU | |
| 15:16:45 | bauzas | mriedem: correct? | |
| 15:17:03 | bauzas | because #2 will verify cell0 with marker #1 and we'll pass on it | |
| 15:17:41 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Plumbing for tagged nic attachment https://review.openstack.org/415622 | |
| 15:17:43 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Delete nic metadata when detaching interface https://review.openstack.org/435387 | |
| 15:17:49 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Delete disk metadata when detaching volume https://review.openstack.org/435082 | |
| 15:17:55 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Plumbing for tagged volume attachment https://review.openstack.org/391941 | |
| 15:17:56 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: API support for tagged device attachment https://review.openstack.org/376021 | |
| 15:19:35 | bauzas | mriedem: nevermind, I'm sure it works | |
| 15:19:46 | bauzas | because we decrease the limit | |
| 15:20:44 | mriedem | bauzas: we wouldn't get #2 if the marker is #1 and it's ACTIVE | |
| 15:20:54 | mriedem | because we process the cells with the ACTIVE instances after cell0 with the ERROR instance | |
| 15:21:14 | bauzas | mriedem: so we would return #1 and #3 ? | |
| 15:21:18 | bauzas | only | |
| 15:21:19 | mriedem | yes | |
| 15:21:19 | bauzas | ? | |
| 15:21:27 | bauzas | mmm, that's another bug, right? | |
| 15:21:28 | mriedem | which is probably also a regression | |
| 15:21:46 | bauzas | lemme double-think on it | |
| 15:21:54 | mriedem | it's not going to get fixed here | |
| 15:21:58 | bauzas | sure | |
| 15:22:02 | mriedem | we search in build requests first, then cell0, then the cells | |
| 15:22:04 | ericyoung | mriedem: regarding https://review.openstack.org/#/c/407440/ (ScaleIO ephemeral) we have a couple of changes to our CI environment and then it should pass | |
| 15:22:12 | mriedem | paging gets wonky | |
| 15:22:13 | ericyoung | hopefully in the nxt day or two | |
| 15:22:15 | mriedem | and is why we said searchlight | |
| 15:22:23 | mriedem | ericyoung: ok | |
| 15:23:07 | bauzas | mriedem: well, so we should only set the marker to the last instance UUID we have, which can be either in cell0 or even a BuildRequesrt | |
| 15:23:32 | bauzas | something like, try to get all you can do with the same marker, and then only set the marker to the last ID you got | |
| 15:24:05 | bauzas | mriedem: in that case, I'd recommend to modify https://review.openstack.org/#/c/468559/ to tackle all those issues | |
| 15:24:38 | mriedem | bauzas: if your marker is on #1 then #2 wouldn't be in the page anyway | |
| 15:24:57 | mriedem | because #1 is ACTIVE | |
| 15:24:59 | mriedem | which means it's in a cell | |
| 15:25:25 | mriedem | our pages right now are basically scheduling (build requests), error (cell0) and then everything else | |
| 15:25:35 | bauzas | that I know | |