Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-05-18
16:03:50 mriedem sean-k-mooney: i see the problem
16:03:50 dansmith sfinucan: cool
16:03:51 mriedem https://github.com/openstack-dev/devstack/blob/stable/ocata/lib/tempest
16:03:57 mriedem the devstack flag didn't get into devstack stable/ocata
16:04:05 mriedem and tempest.conf is configured on the old side in the grenade run
16:04:15 sfinucan dansmith: ...however, 'nova.tests.unit.virt.test_hardware.HelperMethodsTestCase.test_dict_numa_topology_to_obj' is failing for the other two and I can't figure out why
16:04:37 sean-k-mooney mriedem: ah ok that makes sense
16:04:39 sfinucan If you have a chance to look at that at some point, could you?
16:04:53 sfinucan For some reason, a dict is not behaving like a dict :S
16:05:11 sfinucan stupid ovo
16:05:41 cfriesen kashyap: yes, I think that's okay. the implicit all-cpus was just in the nova code.
16:05:52 mriedem sean-k-mooney: so i just needed to backport the fix to devstack https://review.openstack.org/#/c/466033/
16:06:07 mriedem now having done that, the grenade live migration job is still broken, as discussed with sdague earlier
16:06:25 mriedem and i'm not sure i have the energy atm to untangle that mess
16:07:06 sean-k-mooney mriedem: yes backport or swap to useing the test-config update to the file which is supported on ocata already https://github.com/openstack-dev/devstack/blob/stable/ocata/stack.sh#L1405
16:07:14 kashyap cfriesen: True. Okay, I'll go cook food, and head to my Dutch class, then :-)
16:08:57 sean-k-mooney mriedem: oh the flag is new. backporting makes sense then
16:13:03 sean-k-mooney sfinucan: how dare you disbarage ovo :P
16:13:53 sfinucan sean-k-mooney: I'm sure it's lovely, but I've just burnt through a full day looking at ovo-related issues. I ain't feeling it today ;)
16:14:49 sean-k-mooney there a child of nova object you have to treat them kindly but ya they can be a pain.
16:16:04 alex_xu mriedem: sdague sfinucan oomichi there are some nova bug fix in the etherpad from bugsmash https://etherpad.openstack.org/p/OpenStack-Bug-Smash-Pike-Suzhou-Bug-List, appreciate if you can review few of them when you have free time, thanks in advance!
16:17:03 openstackgerrit Alex Xu proposed openstack/nova master: Correct the error message for query parameter validation https://review.openstack.org/465785
16:19:35 dansmith mriedem: so, I'm getting this on my cells test: http://logs.openstack.org/37/458537/20/check/gate-tempest-dsvm-multinode-live-migration-ubuntu-xenial/ab364da/logs/testr_results.html.gz
16:19:48 dansmith mriedem: but I'm failing because of lack of shared storage, which shouldn't be affected by what I'm doing
16:19:54 dansmith oh, um, hmm..
16:20:10 dansmith I wonder if it's because I'm mucking with the config for compute..
16:23:17 dansmith mriedem: isn't there some late hooky stuff going on from the nova tree for the live migration job?
16:23:18 dansmith sdague: ^
16:23:18 openstackgerrit Stephen Finucane proposed openstack/nova master: nova-manage: Deprecate 'log' commands https://review.openstack.org/453176
16:23:18 openstackgerrit Stephen Finucane proposed openstack/nova master: nova-manage: Deprecate 'host' commands https://review.openstack.org/453175
16:23:19 openstackgerrit Stephen Finucane proposed openstack/nova master: nova-manage: Deprecate 'project', 'account' commands https://review.openstack.org/453174
16:23:42 sfinucan alex_xu: I'll trade you some of yours for some of mine ;) ^^^
16:24:05 sfinucan artom: Addressed your comment in above change, if you're about ^
16:24:26 zzzeek sdague: what is a "5 byte utf8" character ?
16:25:18 artom sfinucan, I'm about
16:28:41 sfinucan zzzeek: UTF-8 chars can vary in size from 1 to 6 bytes, it appears https://stackoverflow.com/q/5290182/
16:28:47 sdague zzzeek: there was a nova bug on it at some point
16:29:01 zzzeek sdague sfinucan OK but unicode itself only needs four bytes
16:29:15 zzzeek sdague sfinucan MySQL / Posgresql / anything only goes out to 4 bytes max
16:29:36 sdague ok, good to know, it was in the realm of "um... so not able to go there"
16:29:46 zzzeek https://en.wikipedia.org/wiki/UTF-8#Description
16:29:47 zzzeek OK
16:30:07 sean-k-mooney sfinucan: i thinkg its 1-4 bytes if i remember correctly
16:30:11 zzzeek sdague: thougth that was weird, b.c. why not refer to "6 byte support" :)
16:30:18 sdague I guess I had misunderstood that unicode consortium had crossed that line
16:30:40 sfinucan sean-k-mooney: Are you saying Google lied to me? :O
16:30:41 mdbooth IIRC UTF-8 characters can be arbitrarily many bytes, because they may combine arbitrarily many codepoints
16:31:23 zzzeek mdbooth: the hypotthetical max is 6 bytes
16:31:29 sfinucan And this is the point where I go chuck a ball about in the sunshine :)
16:31:30 sean-k-mooney sfinucan: yes hehe ill double check but when i was working with it in the past i rember it being 4 bytes max for some reason
16:31:35 openstackgerrit Dan Smith proposed openstack/nova master: WIP: De-complicate some of the instance delete path https://review.openstack.org/458636
16:31:36 openstackgerrit Dan Smith proposed openstack/nova master: Move to proper target_cell calling convention https://review.openstack.org/458634
16:31:37 openstackgerrit Dan Smith proposed openstack/nova master: Make target_cell() yield a new context https://review.openstack.org/458635
16:31:38 openstackgerrit Dan Smith proposed openstack/nova master: WIP: Debug race in delete-while-booting test https://review.openstack.org/461513
16:31:39 mdbooth zzzeek: I think that's for a single codepoint?
16:31:40 openstackgerrit Dan Smith proposed openstack/nova master: DNM: Canary test against multi-cell devstack https://review.openstack.org/458537
16:31:40 openstackgerrit Dan Smith proposed openstack/nova master: Target cell on local delete https://review.openstack.org/464059
16:31:56 mdbooth A character may be multiple codepoints
16:32:08 zzzeek mdbooth: now that you have to show me
16:32:09 zzzeek :)
16:32:21 zzzeek mdbooth: im pretty sure no such voodoo is used in practice
16:32:35 mdbooth zzzeek: Like a-umlaut, for eg
16:32:41 mdbooth IIRC that's 2 code points, but 1 char
16:32:47 zzzeek mdbooth: i mean, in databases
16:33:14 mdbooth Well it's all strings ultimately.
16:33:27 mdbooth It boils down to the definition of a character.
16:34:03 dansmith please don't let my snoring interrupt this riveting conversation
16:34:09 zzzeek mdbooth: OK! well. writing an email here, specifics, things that affect us right now :)
16:35:03 sdague zzzeek: I'll update the pg spec with that, it's good info
16:39:14 mdbooth dansmith zzzeek: Heh, yeah, its unlikely to be of interest to us :) It's of interest when you need to display text and, for eg, select characters with a mouse. I read a fantastic article about it recently but I can't find it right now.
16:39:15 sean-k-mooney sfinucan: looking at https://tools.ietf.org/html/rfc3629 section 10 "Security Considerations" we are both right.... the utf-8 encounding would support up to 6 bytes but such encounding produce illegal and only up to 4 bytes are vailid
16:40:05 mdbooth Questions like: what should pressing backspace do in a text editor are actually really hard to answer.
16:47:22 sean-k-mooney sdague: zzzeek mdbooth if ye just define the field as an NVARCHAR in sqlalchemy and set the coalation type for the table to utf8_bin or utf8_general_ci/utf8_unicode_ci the db should handel it its self
16:50:47 openstackgerrit Jianghua Wang proposed openstack/nova master: Skip policy rules on attach_network for none network allocation https://review.openstack.org/465903
16:55:31 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/nova master: Read Neutron port 'binding_profile' during boot https://review.openstack.org/449257
16:55:31 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/nova master: Add Neutron port capabilities to devspec in request https://review.openstack.org/451777
16:55:32 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/nova master: Format NIC features using os-traits definitions https://review.openstack.org/466051
16:57:27 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/nova master: Read Neutron port 'binding_profile' during boot https://review.openstack.org/449257
18:17:20 openstackgerrit melanie witt proposed openstack/nova master: Cache database and message queue connection objects https://review.openstack.org/465042
18:18:27 ZZelle_ sfinucan, hi
18:34:09 marst edleafe: are you around?
18:36:45 openstackgerrit christopher.uhler proposed openstack/nova master: Libvirt volume driver for Veritas HyperScale https://review.openstack.org/443951
18:37:09 openstackgerrit Dan Smith proposed openstack/nova master: DNM: Canary test against multi-cell devstack https://review.openstack.org/458537
18:53:44 cdent marst: edleafe is probably on a plane on his way to pycon
18:59:08 marst cdent: thanks! I've just read his blog post
19:15:18 openstackgerrit Sean Dague proposed openstack/nova master: see if we can use pg wheels with no bindep https://review.openstack.org/466081
19:26:09 mriedem dansmith: did you figure out http://logs.openstack.org/37/458537/20/check/gate-tempest-dsvm-multinode-live-migration-ubuntu-xenial/ab364da/logs/testr_results.html.gz ?
19:47:47 openstackgerrit Cedric Brandily proposed openstack/nova master: Correct _ensure_console_log_for_instance implementation https://review.openstack.org/466088
19:48:56 mriedem this seems wrong http://logs.openstack.org/42/465042/6/check/gate-nova-tox-functional-py35-ubuntu-xenial/9829a45/console.html#_2017-05-18_18_43_03_020297
19:49:00 mriedem b"KeyError: 'allocations'"
19:49:56 dansmith mriedem: I found the hook yeah, still working on a fix
19:50:07 mriedem ok,
19:50:21 mriedem i'm working on disabling the grenade live migration job from running on pike until we fix it, or queens
19:50:28 mriedem since right now it's just burning nodes
19:51:53 mriedem this is very weird - http://logs.openstack.org/42/465042/6/check/gate-nova-tox-functional-py35-ubuntu-xenial/9829a45/console.html#_2017-05-18_18_43_03_020297 it's like in the middle of a functional test, we got a different response from placement
19:52:07 mriedem like it's a GET / response
19:52:21 mriedem but was expecting GET /resource_providers/d41180aa-047e-4bc3-b0f3-11532683400d/allocations
19:54:44 openstackgerrit Cedric Brandily proposed openstack/nova master: Correct _ensure_console_log_for_instance implementation https://review.openstack.org/466088
19:58:31 melwitt mriedem: yeah, I didn't know what to make of that
20:03:21 mriedem could have something to do with running those services under eventlet in our functional tests?

Earlier   Later