| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-06-13 | |||
| 11:37:47 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Add BDM to InstancePayload https://review.openstack.org/448779 | |
| 11:50:10 | openstackgerrit | OpenStack Proposal Bot proposed openstack/nova master: Updated from global requirements https://review.openstack.org/472814 | |
| 12:33:23 | bauzas | johnthetubaguy: sure, will look | |
| 12:57:20 | cdent | edleafe: are you planning to push to https://review.openstack.org/#/c/473627/ soon? if so I'll wait to review. your responses to dtantsur suggest yes-ish | |
| 13:02:50 | openstackgerrit | Merged openstack/nova master: Add bool_from_string for force-down action https://review.openstack.org/473564 | |
| 13:03:17 | mordred | cdent: I have found two new and interesting tidbits that don't seem to be reflected in nova's api docs | |
| 13:03:36 | cdent | mordred: I'd expect nothing less from you | |
| 13:03:40 | mordred | cdent: the first is that if you're booting from volume, you apparently need to POST to /os-volumes_boot instead of /servers (at least, that's what novaclient does) | |
| 13:04:15 | sdague | mordred: that doesn't seem right | |
| 13:04:25 | mordred | cdent: the second is that if you are booting from an image but attaching volumes at boot via block_device_mapping, you need to pass the image in the block_device_mapping wiht a destination_type of local: | |
| 13:04:27 | mordred | https://github.com/openstack/python-novaclient/blob/master/novaclient/v2/servers.py#L737-L745 | |
| 13:04:56 | cdent | mordred: you've used the two red flag terms: "boot from volume" and "bdm" | |
| 13:05:03 | mordred | sdague: https://github.com/openstack/python-novaclient/blob/master/novaclient/v2/servers.py#L1424-L1431 | |
| 13:05:20 | mordred | sdague: if that's not needed, I'll happily change it in shade and send a patch for novaclient | |
| 13:05:38 | sdague | mordred: https://github.com/openstack/nova/blob/d3de541784b094e32cf66e1759fc73781f031607/nova/api/openstack/compute/routes.py#L303-L304 | |
| 13:05:49 | sdague | yeh, it looks like it's a route that was merged a long time ago | |
| 13:05:54 | mordred | awesome | |
| 13:05:59 | mordred | I'll send a novaclient patch | |
| 13:06:47 | sdague | that warrents a pickaxe journey | |
| 13:07:01 | edleafe | cdent: yeah, just running tests before pushing | |
| 13:07:03 | cdent | don't forget your headlamp | |
| 13:07:10 | cdent | edleafe: ✔ | |
| 13:08:15 | cdent | mordred: so the concern on the second point is that there's some magical fiddling that's being done in novaclient (and is required to get proper behavior) but the docs don't say so? | |
| 13:08:48 | openstackgerrit | Monty Taylor proposed openstack/python-novaclient master: Stop posting to os-volumes_boot https://review.openstack.org/473833 | |
| 13:09:23 | mordred | cdent: yes. it's not bad behavior - in fact, it makes total sense | |
| 13:09:31 | mordred | but it was surprising to me | |
| 13:09:40 | mordred | I'm not sure the BEST way to document that though | |
| 13:09:49 | openstackgerrit | Ed Leafe proposed openstack/nova master: Extract custom resource classes from flavors https://review.openstack.org/473627 | |
| 13:10:04 | edleafe | cdent: dtantsur: ^^ | |
| 13:10:22 | dtantsur | thanks! will get back to it after the meeting(s) | |
| 13:10:39 | cdent | mordred: start with BETTER and then maybe BEST will reveal itself? | |
| 13:11:21 | mordred | cdent: you're trying to entice me to document block_device_mapping comprehensively, I can feel it | |
| 13:11:48 | cdent | I wouldn't want to get in the way of your feelings | |
| 13:11:53 | sdague | mordred: it definitely started off as a whole different extension - https://github.com/openstack/nova/commit/97503418bea47ca38d2181e5dc5fa710d6e15df3 | |
| 13:12:05 | sdague | I'm still having trouble figuring out where it merged | |
| 13:12:56 | openstackgerrit | Feodor Tersin proposed openstack/nova master: POC: Imagebackend - on cold migration https://review.openstack.org/473835 | |
| 13:13:35 | ftersin | mdbooth: hi. i tried to refactor that code partially. ^^^^ | |
| 13:15:04 | ftersin | mdbooth: but... this is still inconsistent, does not look correct, and requires much more refactoring | |
| 13:16:41 | ftersin | mdbooth: so... i have to admit that i've gave up. i do not know how to do there 'not a big one' refactoring | |
| 13:17:33 | ftersin | mdbooth: if you have other idea, pls, suggest me smth | |
| 13:17:56 | openstackgerrit | OpenStack Proposal Bot proposed openstack/nova master: Updated from global requirements https://review.openstack.org/472814 | |
| 13:17:57 | mdbooth | ftersin: In a meeting, will try to look today. | |
| 13:19:11 | ftersin | mdbooth: ok, thx | |
| 13:21:44 | openstackgerrit | Feodor Tersin proposed openstack/nova master: Implement ScaleIO image backend https://review.openstack.org/407440 | |
| 13:27:37 | mriedem | efried: is this done now, the services tokens series https://review.openstack.org/#/q/topic:bp/use-service-tokens | |
| 13:27:41 | mriedem | all that's left is dnm testing patches | |
| 13:27:57 | efried | mriedem I believe it is, but was hoping someone else could verify | |
| 13:28:16 | efried | Put another way, you beat me to the question ;-) | |
| 13:29:12 | mriedem | based on the work items in the spec https://review.openstack.org/#/c/426933/5/specs/pike/approved/use-service-tokens.rst@166 i'd say it's done | |
| 13:30:26 | mriedem | actually | |
| 13:30:41 | mriedem | do we actually send the service user tokens with glance requestes? | |
| 13:30:46 | mriedem | *requests | |
| 13:31:50 | efried | mriedem I believe by virtue of https://review.openstack.org/#/c/419662/ | |
| 13:32:10 | mriedem | ah, the topic was different so i missed that one | |
| 13:32:21 | efried | Yeah. | |
| 13:32:50 | efried | mriedem I'm not going to be able to abandon the DNMs, cause I don't own 'em. Is that something you have the power to do? | |
| 13:32:56 | mriedem | i can do that | |
| 13:33:09 | efried | Cool cool. | |
| 13:34:59 | mriedem | sdague: want to abandon this? https://review.openstack.org/#/c/416733/ | |
| 13:35:06 | sdague | mriedem: looking | |
| 13:35:37 | sdague | sure | |
| 13:35:47 | sdague | did we run a test to make sure this was all working as expected? | |
| 13:36:01 | mriedem | sdague: it's enabled in the nova-next job | |
| 13:36:14 | sdague | mriedem: right, but unless you tune down the timeouts | |
| 13:36:20 | sdague | you aren't going to hit it in a gate job | |
| 13:36:31 | mriedem | that's what the abandoned nova dnm patches were for | |
| 13:36:40 | sdague | ok, sure | |
| 13:36:41 | mriedem | they enabled it and did a long sleep during something like live migration | |
| 13:36:53 | sdague | ok, sounds good | |
| 13:47:59 | openstackgerrit | Béla Vancsics proposed openstack/nova master: Reduced the complexity of the _init_instance method https://review.openstack.org/374661 | |
| 14:04:01 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Use plain routes list for user_data instead of stevedore https://review.openstack.org/462183 | |
| 14:04:08 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Use plain routes list for os-server-groups endpoint instead of stevedore https://review.openstack.org/462169 | |
| 14:06:53 | openstackgerrit | Mathieu Gagné proposed openstack/nova master: Add ability to signal and perform online volume size change https://review.openstack.org/454322 | |
| 14:10:14 | mriedem | bauzas: the change on master is merged for https://review.openstack.org/#/c/472687/ if you want to get that in, then we can do an ocata release | |
| 14:10:37 | bauzas | mriedem: easy peasy | |
| 14:28:08 | openstackgerrit | Eric Fried proposed openstack/nova master: Glance download: only fsync files https://review.openstack.org/451094 | |
| 14:29:44 | dansmith | sdague: so this is passing a canary from the nova side: https://review.openstack.org/#/c/436094/ | |
| 14:29:57 | dansmith | sdague: would appreciate it if you'd take another pass through that to see if we can get that landed soonlyish | |
| 14:38:03 | edmondsw | mriedem sdague I think one of you had reminded us about adding powervm to the support matrix... we've got that change up for review: https://review.openstack.org/#/c/470999/ | |
| 14:39:02 | mriedem | yeah i did, just need to get some free time | |
| 14:39:04 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Handle uuid in HostAPI.compute_node_get https://review.openstack.org/461532 | |
| 14:39:04 | openstackgerrit | Matt Riedemann proposed openstack/nova master: WIP: Handle uuids in os-hypervisors API https://review.openstack.org/471899 | |
| 14:39:05 | openstackgerrit | Matt Riedemann proposed openstack/nova master: WIP: Use uuid for id in Services API https://review.openstack.org/464280 | |
| 14:40:07 | sdague | dansmith: the grenade bit is already in? (i can't remember) | |
| 14:40:15 | dansmith | sdague: yar | |
| 14:40:35 | dansmith | sdague: no depends-on for that patch, and just a canary that depends-on it for testing | |
| 14:40:44 | dansmith | sdague: https://review.openstack.org/#/c/473565/ | |
| 14:40:47 | dansmith | canary ^ | |
| 14:41:34 | dansmith | sdague: if you think we need a canary against any other project let me know.. I don't really know what other sorts of complicated jobs we have in other places that might be affected | |
| 14:41:49 | sdague | I think it's probably fine | |
| 14:42:02 | dansmith | cool | |
| 14:42:27 | openstackgerrit | Béla Vancsics proposed openstack/nova master: Reduced the complexity of the _init_instance method (2) https://review.openstack.org/420619 | |
| 14:42:58 | mriedem | dansmith: i see an issue in the ironic tinyipa job in there http://logs.openstack.org/94/436094/35/check/gate-tempest-dsvm-ironic-ipa-wholedisk-bios-agent_ipmitool-tinyipa-ubuntu-xenial-nv/1df37be/logs/devstacklog.txt.gz#_2017-06-12_20_26_29_362 | |
| 14:43:03 | mriedem | but that might just already be busted | |
| 14:44:43 | mriedem | actually, | |
| 14:44:52 | mriedem | it only fails that way on that devstack change and the canary patch | |
| 14:44:58 | mriedem | so it's likely an issue from that setup | |
| 14:45:08 | mriedem | http://logstash.openstack.org/#dashboard/file/logstash.json?query=message%3A%5C%22Timed%20out%20waiting%20for%20Nova%20hypervisor-stats%20count%20%3E%3D%201%5C%22%20AND%20tags%3A%5C%22console%5C%22&from=7d | |
| 14:45:34 | dansmith | ah, yeah, hmm | |
| 14:45:45 | dansmith | host discovery maybe? | |