| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-01-13 | |||
| 16:53:08 | melwitt | yeah, I can't remember how that got decided. probably the argument of "error condition vs not" | |
| 16:53:24 | mriedem | melwitt: the automatic compute node / cell mapping registration would fix this...but it's just not something we've been serious about adding | |
| 16:54:08 | mriedem | thinking about gradually increasing levels of pain, which has been our MO, | |
| 16:54:12 | melwitt | mriedem: yeah. it's not straightforward how to do that because, how do you make assumptions about what cell to auto map new hosts to? usually that's deliberate decision on the part of the operator | |
| 16:54:28 | mriedem | we could do the fallback with a warning in ocata, then make that an error in pike | |
| 16:54:40 | dtp | afk a bit | |
| 16:54:47 | mriedem | melwitt: right you'd need to probably flag some cell mapping as the 'default' to auto-register with | |
| 16:55:19 | mriedem | so in ocata we have the fallback mechanism if we don't find a cell mapping, | |
| 16:55:34 | melwitt | yeah, that's what I had in mind. new stuff goes into some "staging cell" and then after that you have to re-map. though I don't even think we have a way to re-map | |
| 16:55:38 | mriedem | which buys us some time to consider an auto-registration in pike, or remove the fallback and make it an error | |
| 16:56:05 | mriedem | melwitt: we'd probably have to write a REST API for remapping, or nova-manage CLI | |
| 16:56:19 | melwitt | yeah, probably nova-manage | |
| 16:58:46 | mriedem | ok i've added those questions to https://wiki.openstack.org/wiki/Nova-Cells-v2#Open_Questions | |
| 16:58:55 | melwitt | mriedem: I'm wondering if maybe map_cell_and_hosts should stay as returning 0 (since it prints what the deal is, if it found nothing to map) | |
| 16:59:04 | bauzas | mriedem: melwitt: patch is wired, just a few mins for adapting tests | |
| 16:59:54 | melwitt | and later code warnings let you know that you need to do something, if no cell mappings are found while nova is running | |
| 17:00:12 | mriedem | melwitt: _map_cell_and_hosts treats that differently | |
| 17:00:18 | mriedem | if things are already done, it returns a uuid https://github.com/openstack/nova/blob/3ec43d81c324d8229bbb4b5db301175886a049b6/nova/cmd/manage.py#L1275 | |
| 17:00:22 | mriedem | if there are no computes, it returns None | |
| 17:00:36 | mriedem | None is the error, the uuid is the non-error we're already done case | |
| 17:01:26 | melwitt | okay, yeah | |
| 17:01:31 | bauzas | mriedem: melwitt: is the proposed patch http://paste.openstack.org/show/594860/ | |
| 17:02:05 | melwitt | I think we have had this debate before, is doing nothing really an error. I think we usually say it isn't | |
| 17:02:30 | melwitt | from the perspective of returning a non-zero | |
| 17:04:15 | mriedem | it probably depends on the command | |
| 17:08:18 | mriedem | sdague: melwitt: if you're not otherwise distracted, can you take a look at this https://review.openstack.org/#/c/398472/ - would be good to get that in and avoid more rebasing | |
| 17:09:33 | mriedem | bauzas: when you're done with this bug fix are you going to address my comments on the resource providers 'resources' query patch | |
| 17:09:34 | mriedem | ? | |
| 17:09:47 | bauzas | mriedem: I had an hard day | |
| 17:09:53 | bauzas | mriedem: but yeah it's in my pipe | |
| 17:09:59 | mriedem | heh, ok :) | |
| 17:10:33 | mriedem | jay might be able to make those updates too but he's not around right now | |
| 17:12:08 | bauzas | mriedem: no worries, today is not ended for me | |
| 17:12:26 | dims | mriedem : sdague : the py35 change for nova boot/attach/detach/delete working is all green - https://review.openstack.org/#/c/419476/ | |
| 17:12:28 | bauzas | I'm not Jack Bauer but there 24 hours in a day | |
| 17:13:11 | mriedem | sean-k-mooney: if you're willing it'd be great to rebase this series https://review.openstack.org/#/c/410737/ | |
| 17:13:22 | mriedem | the merge conflict is i think b/c of the os-vif unit test fix i made yesterday | |
| 17:14:02 | mdbooth | bauzas: Jack Bauer would say that if he was trying to maintain cover... | |
| 17:14:05 | mdbooth | Just saying | |
| 17:14:06 | sean-k-mooney | mriedem: sure if francesco is not around i have i can rebase them | |
| 17:14:19 | mriedem | sean-k-mooney: i don't know his irc nick, | |
| 17:14:25 | mriedem | or if he's even on irc but don't see him | |
| 17:15:02 | mriedem | frasantoro - nope | |
| 17:15:12 | sean-k-mooney | im just finishing up another commit but ill try and rebase this in the next hour or so if that works for you. | |
| 17:15:31 | mriedem | sounds good, i'll get to it eventually if you can't, it's fine | |
| 17:15:46 | mriedem | i'm going to head to lunch, bbiab | |
| 17:17:34 | openstackgerrit | Sylvain Bauza proposed openstack/nova: Accept simple_cell_setup to partially complete its tasks https://review.openstack.org/420079 | |
| 17:17:56 | bauzas | melwitt: mriedem: ^ | |
| 17:31:46 | dims | thanks sdague ! | |
| 17:42:52 | openstackgerrit | John Garbutt proposed openstack/nova: WIP: Call preallocate_for_instance before scheduler https://review.openstack.org/413124 | |
| 17:42:52 | openstackgerrit | John Garbutt proposed openstack/nova: WIP: Add processed_requests to allocate_for_instance https://review.openstack.org/420103 | |
| 17:52:42 | openstackgerrit | Sarafraj Singh proposed openstack/nova: Adopts keystoneauth with glance client. https://review.openstack.org/412634 | |
| 17:55:26 | cfriesen | any libvirt experts around? When nova calls self._domain.resume(), is that an async call or a sync call? i.e. when it returns has qemu actually transitioned to a "running" state, or does it just send a message but it could be delayed due to scheduling? | |
| 17:59:10 | openstackgerrit | Sarafraj Singh proposed openstack/nova: Adopts keystoneauth with glance client. https://review.openstack.org/412634 | |
| 18:01:40 | openstackgerrit | Sarafraj Singh proposed openstack/nova: Adopts keystoneauth with glance client. https://review.openstack.org/412634 | |
| 18:03:57 | openstackgerrit | John Garbutt proposed openstack/nova: WIP: Call preallocate_for_instance before scheduler https://review.openstack.org/413124 | |
| 18:24:46 | jroll | mriedem1: I'm +1 on the ironic attach/detach/portgroups patches, ruby is out today but I expect she'll be happy with them too | |
| 18:24:59 | jroll | the soft power client patch needs another re-spin | |
| 18:25:13 | jroll | I think that's all I have to update you on... | |
| 18:25:47 | jroll | oh, the hashring thing we did last cycle, that's released in tooz now, do y'all have time for a patch to move to use tooz this cycle yet? (should be mostly ripping code out) | |
| 18:33:37 | mriedem | jroll: umm, maybe | |
| 18:34:14 | mriedem | jroll: there was also the inject nmi patch but i think that was waiting on some dependencies too | |
| 18:34:42 | mriedem | cfriesen: i usually default to think that anything libvirt/qemu does is async | |
| 18:34:45 | jroll | mriedem: yeah, that's not even done in ironic, I don't expect it this cycle | |
| 18:34:49 | mriedem | i think most things are done via an async job | |
| 18:35:01 | mriedem | cfriesen: kashyap or mdbooth or sahid are the people i'd ask | |
| 18:35:06 | mriedem | jroll: ok | |
| 18:36:38 | jroll | mriedem: I'll do the tooz thing now so you can see how simple it is. it passed tests with flying colors about a month ago with some weird depends-ons | |
| 18:37:09 | mriedem | ok, just got back from lunch and still sore from the cells v2 beating i took this morning | |
| 18:37:48 | openstackgerrit | Jim Rollenhagen proposed openstack/nova: Move to tooz hash ring implementation https://review.openstack.org/399022 | |
| 18:37:48 | jroll | oh, easiest merge conflict ever | |
| 18:37:52 | jroll | there you are if curious | |
| 18:38:56 | openstackgerrit | Jim Rollenhagen proposed openstack/nova: Move to tooz hash ring implementation https://review.openstack.org/399022 | |
| 18:39:22 | jroll | much better with the right version | |
| 18:43:31 | openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Ignore me - ry py35 functional with a clean slate https://review.openstack.org/419968 | |
| 18:51:20 | cfriesen | mriedem: thanks...I did some experiments and it looks like it's actually sync...if I send a SIGSTOP to qemu or libvirtd the nova call to self._domain.resume() hangs until I send a SIGCONT | |
| 18:53:26 | openstackgerrit | Dan Peschman proposed openstack/nova: nova-manage cell_v2 map_cell0 exit 0 https://review.openstack.org/420132 | |
| 18:54:33 | dtp | mriedem ^ going to lunch | |
| 18:54:44 | harlowja | dtp thx for helping here :) | |
| 18:55:00 | dtp | why you're welcome harlowja! :) | |
| 18:55:05 | harlowja | :-P | |
| 19:01:12 | sean-k-mooney | mriedem: i have resolved most of the erros in francesco's patches locally but one of his changes need the fastpath os-vif plugin to work which is incorrect http://logs.openstack.org/61/385061/7/check/gate-nova-python27-db-ubuntu-xenial/336f68e/console.html#_2017-01-12_23_22_43_325595 | |
| 19:01:50 | sean-k-mooney | mriedem: im going to invert the order of his patches so https://review.openstack.org/#/c/410737/ can merge | |
| 19:02:07 | sean-k-mooney | or at least be unblocked by the gate | |
| 19:05:40 | openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Ignore me - ry py35 functional with a clean slate https://review.openstack.org/419968 | |
| 19:11:12 | mriedem | sean-k-mooney: sounds good to me | |
| 19:19:17 | mriedem | GET /jaypipes | |
| 19:19:50 | openstackgerrit | sean mooney proposed openstack/nova: os-vif: convert libvirt driver to use os-vif for fast path vhostuser https://review.openstack.org/385061 | |
| 19:19:50 | openstackgerrit | sean mooney proposed openstack/nova: os-vif-util: set vif_name for vhostuser ovs os-vif port https://review.openstack.org/410737 | |
| 19:20:10 | openstackgerrit | Lee Yarwood proposed openstack/nova: WIP - libvirt: workaround findmnt behaviour change https://review.openstack.org/420146 | |
| 19:25:06 | openstackgerrit | Lee Yarwood proposed openstack/nova: WIP - libvirt: workaround findmnt behaviour change https://review.openstack.org/420146 | |
| 19:35:08 | openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Ignore me - ry py35 functional with a clean slate https://review.openstack.org/419968 | |
| 19:35:17 | mriedem | jroll: is there any way for you to tell if this job failed for a known ironic timeout http://logs.openstack.org/73/398473/22/check/gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ipmitool-tinyipa-ubuntu-xenial-nv/3dbaaef/ ? | |
| 19:35:28 | openstackgerrit | Lee Yarwood proposed openstack/nova: libvirt: Use the mirror element to detect when to pivot https://review.openstack.org/399735 | |
| 19:35:36 | mriedem | jroll: b/c it failed on https://review.openstack.org/#/c/398473/ which is moving a bunch of parts around in the compute resource tracker | |
| 19:36:42 | jroll | mriedem: so, this is pretty typical http://logs.openstack.org/73/398473/22/check/gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ipmitool-tinyipa-ubuntu-xenial-nv/3dbaaef/console.html#_2017-01-13_16_02_01_111488 | |
| 19:36:51 | jroll | but could also be because pxe booting got broken somehow or whatever | |
| 19:37:03 | mriedem | it passed on the patch before and after that one so i assume it's not due to the actual change being tested | |
| 19:37:09 | jroll | wtf empty console logs | |
| 19:37:24 | jroll | yeah, probably not then | |