Earlier  
Posted Nick Remark
#openstack-nova - 2017-11-02
13:59:38 jaypipes efried: you didn't push any changes to my refactor series yesterday or this morning did you?
13:59:46 efried jaypipes Nope.
14:00:11 jaypipes k, just checking... finally identified what the damn problem was with the py35 func tests...
14:01:01 efried Dict ordering?
14:01:18 jaypipes nope.
14:01:32 efried jaypipes nova meeting
14:09:00 openstackgerrit Hironori Shiina proposed openstack/nova master: [WIP] Ironic: Get IP address for volume connector from ports/portgroups https://review.openstack.org/517322
14:09:54 mriedem bauzas: you can drop the -2 on this https://review.openstack.org/#/c/491000/
14:21:30 mriedem johnthetubaguy: you'll be in sydney right?
14:36:56 mriedem johnthetubaguy: added the issue to L37 here: https://etherpad.openstack.org/p/sydney-cinder-nova-cp-multi-attach
14:36:58 mriedem ildikov: ^
14:37:47 johnthetubaguy mriedem: oh, good call, I missed that when I reviewed it
14:38:08 ildikov mriedem: thanks
14:38:09 johnthetubaguy mriedem: given the timezone upside down ness, I wasn't going to try remote join
14:38:16 mriedem johnthetubaguy: stvnoyes was the one that actually brought it up, i just wrote the tempest test that verified the issue
14:38:30 mriedem johnthetubaguy: oh so you won't be there, i thought you were for some reason
14:38:31 mriedem alright
14:38:45 johnthetubaguy mriedem: sorry, watching sally's bump
14:38:50 mriedem ah right
14:39:11 mriedem i don't blame you
14:40:08 ildikov mriedem: I will try to sync up with jgriffith before the session
14:40:14 johnthetubaguy could be "marriage limiting" to do something different
14:40:31 ildikov mriedem: as we have options on how to handle this
14:46:36 jaypipes johnthetubaguy: oddly, "watching sally's bump" is a great Hipster band name.
14:46:48 johnthetubaguy jaypipes: heh
14:47:02 jaypipes johnthetubaguy: a band which could have a tremendous Tuba section.
15:01:29 tssurya mriedem : regarding the destroying of request specs records for archived instances ; I remember you telling me to do it similar to how instance_mappings are removed (https://review.openstack.org/#/c/515034/ is the review for removing instance_mappings),
15:02:16 mriedem tssurya: ok
15:02:22 tssurya however, regarding the case of request spec records, as we know there is an open review which destroys the records differently -
15:02:24 tssurya https://review.openstack.org/#/c/391060/
15:02:41 mriedem yeah i don't really agree with sylvain's patch
15:02:44 tssurya I don't know if I should open a new review
15:02:48 tssurya for the same bug
15:03:17 tssurya so then there would be two reviews closing the same bug
15:03:25 mriedem because while he asserts that we recreate reqspecs later in the code if we don't find one, like during evacuate/live migrate/resize, the reqspec we create on the fly is often missing important details, and honestly that was all compat code from newton that is going to be removed anyway
15:03:32 mriedem i'm going to -1 his patch, i'll comment in there
15:03:59 tssurya mriedem : okay thanks! :)
15:04:36 openstackgerrit Balazs Gibizer proposed openstack/nova master: Factor out duplicated notification sample data (2) https://review.openstack.org/452819
15:04:36 openstackgerrit Balazs Gibizer proposed openstack/nova master: Factor out duplicated notification sample data https://review.openstack.org/452818
15:04:37 openstackgerrit Balazs Gibizer proposed openstack/nova master: Deduplicate instance.power-off notification samples https://review.openstack.org/475860
15:04:37 openstackgerrit Balazs Gibizer proposed openstack/nova master: Deduplicated instance.(un)pause notification samples https://review.openstack.org/452820
15:04:38 openstackgerrit Balazs Gibizer proposed openstack/nova master: Deduplicate instance.live_migration notification samples https://review.openstack.org/515745
15:04:38 openstackgerrit Balazs Gibizer proposed openstack/nova master: Deduplicate instance.interface_attach samples https://review.openstack.org/515740
15:04:39 openstackgerrit Balazs Gibizer proposed openstack/nova master: Deduplicate instance.reboot notification samples https://review.openstack.org/515751
15:07:30 mriedem done
15:08:00 openstackgerrit Ed Leafe proposed openstack/nova master: Add Selection objects https://review.openstack.org/499239
15:08:01 openstackgerrit Ed Leafe proposed openstack/nova master: Modify select_destinations() to return objects and alts https://review.openstack.org/510159
15:08:01 openstackgerrit Ed Leafe proposed openstack/nova master: Return Selection objects from the scheduler driver https://review.openstack.org/495854
15:08:02 openstackgerrit Ed Leafe proposed openstack/nova master: Move the claim_resources method to scheduler utils https://review.openstack.org/511357
15:08:02 openstackgerrit Ed Leafe proposed openstack/nova master: Change RPC for select_destinations() https://review.openstack.org/516707
15:08:03 openstackgerrit Ed Leafe proposed openstack/nova master: Make conductor pass and use host_lists https://review.openstack.org/511358
15:08:13 tssurya mriedem : ack
15:10:34 openstackgerrit Matt Riedemann proposed openstack/nova master: Remove instance.keypairs migration code https://review.openstack.org/517158
15:13:16 edleafe mriedem: The legacy_find_hosts fix you have is something I fixed in a later patch in my series. https://review.openstack.org/#/c/510159/22/nova/scheduler/filter_scheduler.py@365
15:13:37 edleafe mriedem: how about I move that fix earlier in the series, so that your fix isn't needed?
15:15:56 edleafe the problem is that the drivers shouldn't be messing with the return type. That should be done by the manager at the RPC boundary
15:24:24 mriedem edleafe: i think it should be fixed outside of your series
15:24:36 mriedem because it's busted for anyone picking this up today from master
15:25:02 mriedem this isn't something that hits the manager either, it's an IndexError within the FilterScheduler itself
15:27:01 edleafe mriedem: yeah, I know the problem. So you want me to rebase on top of your fix?
15:27:13 mriedem edleafe: sure
15:28:30 edleafe ok, then how about I update your fix with the simpler version I added later?
15:29:14 openstackgerrit Balazs Gibizer proposed openstack/nova master: Document the real behavior of notify_on_state_change https://review.openstack.org/516264
15:29:28 edleafe _get_alternate_hosts() won't add alternates, but will make the right structure, if return_alternates is False
15:29:33 mriedem edleafe: i don't think that's necessary - my fix is contained to that one method, you can still refactor this to remove the include_alternatives parameter later in your series
15:29:56 mriedem assuming you're talking about this https://review.openstack.org/#/c/510159/22/nova/scheduler/filter_scheduler.py@176
15:30:21 mriedem how about just rebase on top of my fix and resolve the merge conflict, which should be tiny
15:31:42 edleafe mriedem: sure. Just seemed odd to fix the same thing twice
15:34:51 openstackgerrit Matt Riedemann proposed openstack/nova master: Implement power_off/power_on for the FakeDriver https://review.openstack.org/509935
15:39:42 mriedem gcb_: replied in https://bugs.launchpad.net/oslo.log/+bug/1729475 and https://review.openstack.org/#/c/478285/1/oslo_log/formatters.py
15:39:43 openstack Launchpad bug 1729475 in oslo.log "Exception traces are logged when warning() is used even if exc_info=False is specified" [Undecided,New]
15:39:51 mriedem gcb_: is this exception summary stuff all new since ocata?
15:40:12 mriedem if the caller is passing exc_info=False, it seems oslo.log should respect that regardless of the warning() log level and not dump a trace
15:41:15 mriedem otherwise i guess we have to change all of these warnings to info log level
15:45:00 jaypipes ggVGd
15:45:09 jaypipes oh ffs, vim fail
15:46:11 efried jaypipes You gonna rebase the nrp series to pick up https://review.openstack.org/516662 ?
15:46:23 edleafe too bad 'ffs' isn't a vim command
15:46:33 jaypipes edleafe: indeed :)
15:46:59 jaypipes efried: as soon as I finish the refactor thing in about 3 mins
15:47:44 openstackgerrit Jianghua Wang proposed openstack/nova master: XenAPI: get vGPU stats from hypervisor https://review.openstack.org/512965
15:50:26 openstackgerrit Jay Pipes proposed openstack/nova master: build ProviderSummary objects in sep function https://review.openstack.org/516779
15:50:26 openstackgerrit Jay Pipes proposed openstack/nova master: begin refactor AllocCandidates._get_by_filters() https://review.openstack.org/516778
15:50:27 openstackgerrit Jay Pipes proposed openstack/nova master: create allocation request for single provider https://review.openstack.org/516781
15:50:27 openstackgerrit Jay Pipes proposed openstack/nova master: build alloc request resources for shared resources https://review.openstack.org/516780
15:50:28 openstackgerrit Jay Pipes proposed openstack/nova master: required traits for no sharing providers https://review.openstack.org/517027
15:50:28 openstackgerrit Jay Pipes proposed openstack/nova master: finish refactor AllocCandidates._get_by_filters() https://review.openstack.org/516782
15:50:29 openstackgerrit Jay Pipes proposed openstack/nova master: WIP: handle traits with sharing providers https://review.openstack.org/517119
15:50:47 jaypipes efried: ^^ addressed all of your comments (I think)
15:50:53 efried jaypipes Cool man.
15:51:10 jaypipes efried: reduced the complexity of the sharing code with itertools.product()...
15:51:14 jaypipes efried: see inline.
15:51:54 jaypipes efried: the only comment I didn't respond to was the request for "lots more permutations" of functional tests. I'm not going to try and add every test permutation in a single patch.
15:51:57 efried jaypipes Instead of zip? That's good. Might even fix that one bug alex_xu was looking at.
15:52:35 efried jaypipes I was thinking to write a bunch of tests over the next week or so.
15:52:51 jaypipes efried: that would be great.
15:53:03 efried Because whereas this stuff is pretty tough to code, it's fairly straightforward to reason about how the API should respond.
15:53:15 efried ...in most cases :)
15:53:18 jaypipes efried: it's just I added a shit-load of new tests in the traits patches and would rather add separate patches for ever-more-complicated test scenarios.
15:54:52 kashyap mdbooth: A question - During an instance boot, Glance image is copied to the local host image cache. Then, Nova creates a qcow2 overlay with this copied image as a backing file. Now, does anything modify this backing file?

Earlier   Later