| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-11-28 | |||
| 23:02:27 | mordred | I mean - efried will probably beat me to it - but getting rid of that catalog_info parameter at least is a step in teh right direction | |
| 23:02:31 | mordred | efried: yup! | |
| 23:02:41 | sorrison | mriedem: re cinder catalog in nova.conf yes the only way we can get it to work is by setting endpoint_template | |
| 23:02:54 | efried | You ought to be able to a) name your service any of the list of valid things, and b) search for the endpoint using any of that same list of valid things. | |
| 23:03:18 | mordred | yup. and you should absolutely be able to omit service_name which should default to None | |
| 23:03:29 | jaypipes | efried: unless I'm totally bonkers... but AFAIK, when Ironic is in the mix, that ProviderTree can contain thousands of root providers. | |
| 23:03:50 | efried | and yeah, long-standing TODO to get rid of all that bizarro client construction gorp | |
| 23:04:04 | mordred | and you should not only be able to omit service-type- you should really always omit service-type because using different service types is CRAZY - but you should be able to configure it anyway | |
| 23:04:24 | mordred | efried: I started looking at all that before the summit, then had thanksgiving | |
| 23:04:39 | mordred | I'd love to make a patch for y'all with a strawman of ripping a ton of gorp out | |
| 23:05:12 | efried | I'll be happy to review same, but I'm not likely to have time to code it up myself, alas. | |
| 23:05:27 | efried | I'm up to here | |
| 23:06:06 | mordred | cool. well - you did the last one - so it's my turn this time I think | |
| 23:06:20 | mriedem | efried: remember you had https://review.openstack.org/#/c/508345/ | |
| 23:06:24 | mriedem | but very out of date by now | |
| 23:06:42 | efried | oo, indeed | |
| 23:06:50 | mordred | mriedem: yeah - I'll likely start by seeing if I can update that :) | |
| 23:08:11 | efried | jaypipes: Hum, I see your point. I did have it rigged at some point to just invalidate the tree of the failing provider. That turned out to be nontrivial and hacky, which is why I went the "invalidate everything" route. But yeah, I may have to restore that logic :( | |
| 23:16:04 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Make [cinder]/catalog_info no longer require a service_name https://review.openstack.org/620738 | |
| 23:16:05 | mriedem | sorrison: see how ^ floats your boat | |
| 23:17:27 | sorrison | mriedm: looks very buoyant, thanks! | |
| 23:20:19 | sorrison | Does anyone about any big openstack installs that connect to rabbitMQ with SSL, we've been having a lot of fun with https://bugs.launchpad.net/oslo.messaging/+bug/1800957/ so far it seems use of SSL for rabbit is very low | |
| 23:20:19 | openstack | Launchpad bug 1800957 in oslo.messaging "Upgrading to pike version causes rabbit timeouts with ssl" [Undecided,Incomplete] - Assigned to Ken Giusti (kgiusti) | |
| 23:28:32 | mriedem | sorrison: i seem to remember klindgren saying something similar a long time ago | |
| 23:28:44 | mriedem | so maybe SpamapS has an idea? | |
| 23:33:46 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Make [cinder]/catalog_info no longer require a service_name https://review.openstack.org/620738 | |
| 23:37:12 | mriedem | cfriesen: jackding: fyi hpet related https://bugs.launchpad.net/nova/+bug/1805087 | |
| 23:37:12 | openstack | Launchpad bug 1805087 in OpenStack Compute (nova) "libvirt+KVM: High CPU usage on Windows 10 (1803) guests" [Undecided,New] | |
| 23:38:31 | mriedem | turns out it's not just windriver OS images that need hpet | |
| 23:42:23 | SpamapS | sorrison: mriedem yes GoDaddy does use TLS for RabbitMQ. I don't work there anymore though. | |
| 23:42:55 | SpamapS | With cells v1, and cells of up to 1000 hv's. | |
| 23:43:02 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Note the aggregate allocation ratio restriction in scheduler docs https://review.openstack.org/620713 | |
| 23:43:39 | sorrison | SpamapS: thanks. I'll send klindgren an email and get some more info off him | |
| 23:43:53 | SpamapS | For cells v2 the TLS should still be in use, but the cells v2 deployment wasn't quite scaled out when I was there. | |
| 23:44:19 | SpamapS | sorrison: out of curiosity, what issue are you seeing? | |
| 23:45:36 | sorrison | SpamapS: See https://bugs.launchpad.net/oslo.messaging/+bug/1800957/ basically timeouts waiting for replies | |
| 23:45:36 | openstack | Launchpad bug 1800957 in oslo.messaging "Upgrading to pike version causes rabbit timeouts with ssl" [Undecided,Incomplete] - Assigned to Ken Giusti (kgiusti) | |
| 23:45:40 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Provide a useful error message when trying to update non-compute services https://review.openstack.org/620667 | |
| 23:46:49 | sorrison | SpamapS: been having a lot of fun as we didn't see this in testing and I'm still unsure why it happens, but I at least have found a reliable set of versions that work (basically the ocata versions) | |
| 23:47:44 | SpamapS | sorrison: Yeah the big deploy at GoDaddy was still running Liberty when I left about 6 weeks ago. | |
| 23:48:03 | SpamapS | the smaller deployment was Pike with cells v2. | |
| 23:48:27 | SpamapS | RabbitMQ is basically the most hated thing on the OpenStack team at GoDaddy, so, I'm not sure anybody will give you a nuanced answer. ;) | |
| 23:49:27 | sorrison | yeah I am aware of klindgren hate for rabbit, it's been stable for us over the last couple years up until pike upgrade | |
| 23:50:18 | sorrison | we (nectar) have cellsv1, 13 cells ~1000 hypervisors | |
| #openstack-nova - 2018-11-29 | |||
| 00:01:35 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Default zero disk flavor to RULE_ADMIN_API in Stein https://review.openstack.org/603910 | |
| 00:05:06 | SpamapS | IIRC, we had some improvement with newer RabbitMQ versions. | |
| 00:06:04 | sorrison | SpamapS: yeah some of our cells have found upgraded rabbit helps. We tried to upgrade the rabbit that neutron uses and it fell over so we down graded back | |
| 00:36:58 | openstackgerrit | Artom Lifshitz proposed openstack/nova-specs master: Re-propose numa-aware-live-migration spec https://review.openstack.org/599587 | |
| 04:32:01 | openstackgerrit | Ghanshyam Mann proposed openstack/nova stable/queens: Test websocketproxy with TLS in the nova-next job https://review.openstack.org/620776 | |
| 04:44:05 | openstackgerrit | Ghanshyam Mann proposed openstack/nova stable/queens: Move the nova-next job in-tree and update it https://review.openstack.org/620777 | |
| 04:47:09 | openstackgerrit | Ghanshyam Mann proposed openstack/nova stable/queens: Test websocketproxy with TLS in the nova-next job https://review.openstack.org/620776 | |
| 07:24:24 | openstackgerrit | Brin Zhang proposed openstack/nova master: Remove useless sample and add the lack of tests in v266 https://review.openstack.org/614671 | |
| 07:32:24 | openstackgerrit | Brin Zhang proposed openstack/nova master: Remove useless sample tests in v266 https://review.openstack.org/614671 | |
| 09:27:48 | openstackgerrit | Matthew Booth proposed openstack/nova master: Workaround a race initialising version control in db_version() https://review.openstack.org/619622 | |
| 09:33:36 | openstackgerrit | Matthew Booth proposed openstack/nova master: Workaround a race initialising version control in db_version() https://review.openstack.org/619622 | |
| 09:35:42 | openstackgerrit | Jason SUN proposed openstack/nova master: Add task_states when confirm resize https://review.openstack.org/620848 | |
| 09:38:24 | openstackgerrit | Zhenyu Zheng proposed openstack/nova-specs master: Per aggregate scheduling weight (spec) https://review.openstack.org/599308 | |
| 09:40:15 | openstackgerrit | Alvaro Lopez Garcia proposed openstack/nova master: Ensure that periodic reclaim cleans DB deleted instances https://review.openstack.org/323250 | |
| 09:51:24 | ohorecny2_ | Hi #stephenfin, I would like to ask you about your commit https://github.com/openstack/nova/commit/6b3b04d2113915729fd9aced9839338e429e1a01 which implemented PCI NUMA policies, I found that this change is not working properly. So I search about fix and I found this bug: https://bugs.launchpad.net/nova/+bug/1795920 . I would like to ask you if anybody is working on it, or if you have some proposals how to fix it. | |
| 09:51:24 | openstack | Launchpad bug 1795920 in OpenStack Compute (nova) "SR-IOV shared PCI numa not working " [Undecided,Confirmed] | |
| 09:53:26 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Fix misuse of assertTrue https://review.openstack.org/620861 | |
| 10:08:39 | openstackgerrit | Merged openstack/nova master: Remove get_node_uuid https://review.openstack.org/620715 | |
| 10:10:08 | openstackgerrit | Kashyap Chamarthy proposed openstack/nova master: libvirt: Refactor handling of PCIe root ports https://review.openstack.org/620327 | |
| 10:14:45 | openstackgerrit | Alex Xu proposed openstack/nova master: Parse the numa topology from the new flavor when resize instance https://review.openstack.org/620874 | |
| 10:42:47 | mdbooth | bauzas: I need placement to think my compute has more than 2 vcpus in my functional test. Any idea what the best thing to mock would be? | |
| 10:57:44 | mdbooth | bauzas: When you're back, I think I'll be best mocking LibvirtDriver.get_available_resource() | |
| 11:07:28 | bauzas | mdbooth: well, no | |
| 11:07:38 | bauzas | mdbooth: we now have a new method | |
| 11:08:56 | bauzas | mdbooth: but do you need libvirt ? because we have a FakeDriver | |
| 11:10:50 | bauzas | mdbooth: https://github.com/openstack/nova/blob/master/nova/virt/fake.py | |
| 11:11:07 | bauzas | eg. https://github.com/openstack/nova/blob/master/nova/virt/fake.py#L138 | |
| 11:32:59 | openstackgerrit | Ghanshyam Mann proposed openstack/nova stable/queens: Make the nova-next job voting and gating https://review.openstack.org/620888 | |
| 12:10:31 | mdbooth | bauzas: No, I'm specifically testing the LibvirtDriver | |
| 12:10:38 | mdbooth | I *think* I've nearly got it | |
| 12:10:42 | mdbooth | Although.... | |
| 12:11:13 | mdbooth | AAAAAAAAAAAAAAAAAAARGH! methods which secretly modify shared data! | |
| 12:12:23 | sean-k-mooney | mdbooth: dose the method also retrun something? | |
| 12:13:01 | sean-k-mooney | if it just has sideffect its not terrible but if it both retruns things and has sideffect that is a pain | |
| 12:17:58 | mdbooth | bauzas: Sorted now. mocked libvirt methods for fetching disk space, vcpu, and memory, which are used in both places. | |
| 12:24:05 | cdent | mdbooth: tshirt | |
| 12:25:44 | mdbooth | cdent: Hehe | |
| 12:28:06 | mdbooth | cdent: This case wasn't egregious, tbh. LibvirtDriver._get_local_gb_info modified the return from libvirt_utils.get_fs_info() in place, which isn't unreasonable. However, I was mocking get_fs_info(), so it was always returning the same dict. It took me over an hour to work out why my disk space got 1000 times smaller every time it was called. | |
| 12:28:44 | cdent | <- doesn't know haskell | |
| 12:28:47 | mdbooth | FTW! | |
| 12:28:54 | mdbooth | <- neither | |
| 12:29:14 | cdent | of course the number of times I've heard people say that the original authors of nova didn't know python is unfortunately high | |
| 12:29:30 | mdbooth | HAHA | |
| 12:29:31 | cdent | so, you know, maybe it is a goer | |
| 12:30:15 | sean-k-mooney | i didnt know python when i started working on openstck not sure i still dont | |
| 12:31:07 | sean-k-mooney | python is a language that you can pick up as you go along and with the gated trunk/review model we use you learn what you need quickly | |
| 12:31:08 | cdent | yeah, that's the other side of the coin: if you can code, you can usally code anywhere, if you pay attention | |
| 12:31:52 | sean-k-mooney | yep once you lear to corralate different concpets in different languages and how to reason about problems that applies across many languages | |
| 12:32:22 | sean-k-mooney | functional programing people however i have found think differnetly then i do | |
| 12:33:37 | mdbooth | sean-k-mooney: It's a mental switch. It's not as hard to make as you'd think. | |
| 12:34:41 | sean-k-mooney | mdbooth: ya i have tried learning some functional libs in the past and i can use them but its still a style that i reserve for specific uses | |
| 12:34:56 | cdent | there are degrees as well. you can sort of pick and choose what you think matters most. for example: don't be side-effecty | |
| 12:35:21 | mdbooth | That's a general lesson | |
| 12:35:22 | cdent | for collaborative code maintenance I think that one is the most important | |
| 12:35:32 | cdent | a lesson people apparently don't learn often enough :) | |
| 12:36:07 | kashyap | cdent: What makes you think all the current contributors (/looks at self) "know" Python? :D | |