Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-01-26
15:14:02 dansmith mriedem: to drive home that placement config is required
15:14:10 dansmith mriedem: i.e. like anything else, there's not much point in starting if we don't have sufficient config to do anything
15:14:29 claudiub|2 sfinucan: so, VFs / PFs won't show up as assignable PCI devices, rather those are requested from Hyper-V, and Hyper-V will then assign them to the VMs.
15:14:49 sfinucan claudiub|2: as what? VIFs?
15:14:53 johnthetubaguy bauzas: +1 what dansmith said, that sounds right
15:15:15 bauzas johnthetubaguy: dansmith: and it would raise an exception if not CONF.placement?
15:15:33 mriedem ok so the grenade case should be ok b/c newton subnode will be running and ocata compute will have config so will start
15:15:42 dansmith mriedem: right
15:15:43 mriedem but scheduler will use the old RT data because of the newton compute
15:15:47 dansmith yes
15:16:00 claudiub|2 sfinucan: yeah. The vif has a property - if it SR-IOV is enabled for it or not. by default it is false.
15:16:02 dansmith bauzas: yes, or whatever will arrest init_host cleanly
15:16:13 bauzas dansmith: that's my question
15:16:17 bauzas dansmith: should we log and stop ?
15:16:26 bauzas or should we more like "sorry dude"
15:16:36 dansmith I dunno what that means
15:16:40 bauzas heh
15:16:43 bauzas I mean
15:16:45 bauzas two choices
15:16:58 bauzas #1 log an ERROR and return
15:17:07 bauzas #2 raise an exception
15:17:16 bauzas I'd favor #1
15:17:28 dansmith bauzas: that won't stop compute from starting
15:17:38 dansmith that will just leave it further hamstrung
15:17:53 bauzas okay, lemme see how init_host is working
15:18:19 mriedem an unhandled exception from init_host will kill the nova-compute process
15:18:22 mriedem which i think is the goal
15:18:54 mriedem meaning you upgraded the code to ocata but didn't configure it properly,
15:18:57 mriedem so you need to get paged
15:18:59 dansmith right, only an exception will stop it from what I see
15:19:08 bauzas okay, that's my question
15:19:15 bauzas because it's threaded ?
15:19:22 dansmith bauzas: so just raise an exception, which will be ugly, and we could maybe clean that up in service.py after the fact to make it a little less ugly on the eyes
15:19:23 bauzas I mean gt
15:19:30 dansmith nothing related to threading
15:19:33 bauzas anyway
15:19:37 bauzas let's do the exception
15:19:46 bauzas no time for introspecting the caller :)
15:20:31 mriedem i guess https://review.openstack.org/#/c/425524/ didn't work huh
15:21:00 dansmith I didn't look, this conversation was going on when I showed up
15:21:01 dansmith however,
15:21:05 openstackgerrit Gábor Antal proposed openstack/nova: Fixed indentation in virt/libvirt/driver.py https://review.openstack.org/425748
15:21:08 pkoniszewski_ hey, can i have -W/+W there https://review.openstack.org/#/c/407518/ ? jenkins does not want to start gate job for some reason
15:21:10 dansmith the moreI explain it the more I think it's probably the better option anyway
15:21:50 mriedem bauzas: you can remove the dependency on https://review.openstack.org/#/c/425524/ when you respin your nova scheduler change,
15:21:53 mriedem it didn't work anyway
15:22:00 cdent mriedem: I checked into the grenade job and as far as I could tell, the subnode wasn't getting config'd, still
15:22:08 mriedem cdent: yeah it wasn't
15:22:13 mriedem that features yaml is all magic to me
15:22:51 bauzas mriedem: ack
15:23:16 pkoniszewski_ sfinucan: hi, are you around?
15:23:47 sfinucan pkoniszewski_: aye
15:23:57 pkoniszewski_ sfinucan: can i have -W/+W there https://review.openstack.org/#/c/407518/ ? jenkins does not want to start gate job for some reason
15:24:20 openstackgerrit Stephen Finucane proposed openstack/nova: Change exponential function to linear https://review.openstack.org/407518
15:24:35 sfinucan pkoniszewski_: done
15:24:40 pkoniszewski_ sfinucan: thanks!
15:25:32 openstackgerrit Artom Lifshitz proposed openstack/nova: Fix tag attribute disappearing in 2.33 and 2.37 https://review.openstack.org/424759
15:27:00 dansmith I think we probably need to touch the oslo service stuff to make a really graceful exit from service startup
15:27:49 dansmith actually, no
15:28:05 mriedem i don't think so,
15:28:13 mriedem we had a bug where the pci whitelist parser was blowing up,
15:28:14 openstackgerrit Claudiu Belu proposed openstack/nova: Hyper-V PCI Passthrough https://review.openstack.org/420614
15:28:17 mriedem and that just killed the process
15:28:28 mriedem ^ came from the RT on init_host
15:28:35 dansmith mriedem: no I mean to avoid a traceback
15:28:43 mriedem oh
15:29:25 openstackgerrit Dan Smith proposed openstack/nova: WIP: Allow graceful service exits from compute https://review.openstack.org/425751
15:29:43 dansmith mriedem: something like this ^ to clean up what bauzas is adding now .. but later
15:30:03 bauzas dansmith: I'm just writing UTs and then boooom
15:30:13 sfinucan claudiub|2: OK, I'd still like Moshe to jump in but I'm personally happy with that as is
15:30:43 claudiub|2 sfinucan: cool, thanks! :D
15:30:46 bauzas dansmith: you okay if I'm just shamely copying your exception ?
15:30:55 claudiub|2 sfinucan: will try to reach out to him as well, as soon as possible.
15:31:02 bauzas dansmith: mine is too much explicit (PlacementNotConfigured)
15:31:07 sfinucan Side note: we *really* need to find a way to standardized flavor extra specs like we do ImageMetaData
15:31:19 bauzas dansmith: but I was basically doing the same, including the try/except clause in the cmd
15:33:40 mriedem i don't think create() is the correct method to handle,
15:33:46 mriedem that just creates the service object,
15:33:48 dansmith bauzas: just use something else and we can merge my graceful bit later and I will clean it up
15:33:49 mriedem start() is the thing that blows up
15:34:03 bauzas dansmith: okay
15:34:08 dansmith mriedem: yep, you're right
15:34:16 dansmith mriedem: but I think I want to catch around both
15:34:25 dansmith mriedem: in case we had to abort in low-level init or something
15:34:27 mriedem i just don't know if serve() or wait() calls start()
15:34:31 andreykurilin johnthetubaguy: I answered all your comments and all both patches are vaiting for your review:)
15:34:51 dansmith serve calls launch
15:34:54 dansmith which calls start I think
15:35:04 dansmith this is why I made it fail so we can see :)
15:35:17 openstackgerrit Dan Smith proposed openstack/nova: WIP: Allow graceful service exits from compute https://review.openstack.org/425751
15:41:45 cdent mriedem: if we stick to one full circle, I think we should consider that a win
15:42:04 mriedem heh
15:42:12 openstackgerrit melanie witt proposed openstack/nova: Count resources to check quota for cells https://review.openstack.org/416521
15:42:25 mriedem sfinucan: if you're looking for things, https://review.openstack.org/#/c/404990/ and the one after it have +2s
15:42:37 mriedem pretty simple
15:44:41 sfinucan mriedem: I'll give it a shot
15:45:55 mriedem none of that is sqla stuff
15:46:00 mriedem even though it's a jaypipes patch
15:46:00 raj_singh mriedem: If you have time, I got +2's on last two patches of service token bp.
15:47:03 mfeoktistov sfinucan: mriedem: mdbooth: sorry for disturbing you, I updated https://review.openstack.org/#/c/312488/ (rootwrap instead of privsep and small cleanup). Virtuozzo Storage CI fails because we have problem with cinder now, it's not releated to this review.

Earlier   Later