Thursday, 2013-09-05

« Wednesday, 2013-09-04 Index Friday, 2013-09-06 »
*** kgriffs is now known as kgriffs_afk00:01
*** oz_akan_ has joined #openstack-marconi00:30
*** ayoung-afk is now known as admiyo00:31
*** amitgandhi has joined #openstack-marconi00:31
*** nosnos has joined #openstack-marconi00:36
*** amitgandhi has quit IRC01:35
*** amitgandhi has joined #openstack-marconi01:38
*** oz_akan_ has quit IRC02:14
*** oz_akan_ has joined #openstack-marconi02:14
*** oz_akan_ has quit IRC02:19
*** amitgandhi1 has joined #openstack-marconi03:02
*** amitgandhi1 has quit IRC03:04
*** amitgandhi has quit IRC03:05
*** oz_akan_ has joined #openstack-marconi03:25
*** oz_akan_ has quit IRC03:29
*** admiyo has quit IRC03:40
*** flaper87|afk is now known as flaper8707:28
*** flaper87 is now known as flaper87|afk08:04
*** flaper87|afk is now known as flaper8708:18
*** tedross has joined #openstack-marconi11:21
*** nosnos has quit IRC11:24
*** oz_akan_ has joined #openstack-marconi12:12
*** oz_akan_ has quit IRC12:12
*** oz_akan_ has joined #openstack-marconi12:13
openstackgerritA change was merged to stackforge/marconi: fix: Claim can return 404  https://review.openstack.org/4453612:13
flaper87Ha, ccpcabrera is reviewing stuff hidden under a rock. :D12:15
flaper87I know he's somewhere...12:15
oz_akan_:)12:19
openstackgerritFlavio Percoco proposed a change to stackforge/marconi: Make tox use develop instead of sdist  https://review.openstack.org/4522712:59
flaper87openstackgerrit: dude, you're back. You've been in and out the last couple of days. We missed you.13:00
EmilienMflaper87: \o/13:05
flaper87EmilienM: yo!13:07
*** amitgandhi has joined #openstack-marconi13:34
oz_akan_kgriffs_afk: there?14:00
*** kgriffs_afk is now known as kgriffs14:57
oz_akan_kgriffs: hey14:58
kgriffs\o\14:58
kgriffs /o/14:58
kgriffs\o/14:58
oz_akan_\^/14:59
oz_akan_seems like claim patch works fine14:59
* kgriffs and the crowd goes wild!!!15:00
oz_akan_I am not sure about performance but at least there are no more 404s15:00
flaper87<o/ \o> <o/ \o>15:00
kgriffsok, shall I make it official?15:00
oz_akan_yes please15:00
* kgriffs goes off to do that15:00
oz_akan_and I can't wait to test large queue patch on top of that15:00
kgriffsgroovy15:00
kgriffsI'll do that and get my reviewing groove going today so we can reduce the patch backlog15:01
flaper87reviews reviews reviews15:01
flaper87:D15:01
oz_akan_quick guys. I want to have the awesomeness in marconi15:02
amettsflaper87: Regarding https://review.openstack.org/#/c/45227, is virtualenv a requirement for dev environments now?15:03
flaper87ametts: it's always been. Tox depends on it15:04
amettsok15:05
flaper87that patch makes tox use python setup.py develop instead of doing sdist and installing the whole thing in the env15:05
flaper87which means, tox is so freaking fast now15:05
flaper87:D15:05
amettsREADME.rst doesn't mention virtualenv -- should it?15:06
flaper87ametts: naah, that's up to the user15:07
flaper87I mean, whether he wants to install marconi in a venv or globally15:07
amettsWait, what?! :)15:07
flaper87so, running tox is different from testing marconi - as in running it and issuing queries.15:07
flaper87tox is a developer tool15:07
kgriffsinstall_command = pip install -U {opts} {packages}15:07
kgriffscan you comment on that addition to tox.ini ?15:08
amettsOk , so virtualenv is required to run the tests... but not if you want to just run marconi and muck with it.  Or even change code if you don't want to test it. :D15:08
flaper87kgriffs: install_commands allows us to customize the command for installing dependencies. It's being used just because of the -U15:10
flaper87ah, you want me to add the comment to the file15:10
flaper87sorry, lemme do that15:10
openstackgerritFlavio Percoco proposed a change to stackforge/marconi: Make tox use develop instead of sdist  https://review.openstack.org/4522715:12
flaper87kgriffs: done15:12
* flaper87 is still refactoring functional tests15:41
kgriffscool, thanks15:43
kgriffsprotip: another way to speed up pip15:44
kgriffshttps://gist.github.com/kgriffs/639218615:44
kgriffs(TL;DR - use a pypi mirror)15:44
flaper87uu cool, how frequently is crate.io updated?15:44
kgriffsseems to be quite often15:45
flaper87cool15:45
flaper87thx for the tip15:45
kgriffshttps://crate.io/packages/tox/15:45
kgriffsfor example, 1.6.1 is there15:45
kgriffsrock on15:45
flaper87awesome! put it in my pip.conf15:46
kgriffsanybody mind if I just approve this w/o Alejandro's +2 ?15:51
kgriffshttps://review.openstack.org/#/c/45227/15:51
kgriffs(he is on Jury duty, and this is a trivial patch)15:52
flaper87I don't mind!15:52
flaper87:D15:52
flaper87does that count?15:52
flaper87:D15:52
flaper87jokes apart, it's harmless15:52
kgriffssuuuuuuure it does.15:52
kgriffs:D15:52
flaper87I've been using it the whole day15:52
flaper87:D15:52
kgriffscool15:52
* kgriffs does that15:52
flaper87I know run tox just for fun15:52
flaper87:P15:53
kgriffsif it printed pop-tart ascii art, so would I15:53
* kgriffs wants that15:53
openstackgerritA change was merged to stackforge/marconi: Make tox use develop instead of sdist  https://review.openstack.org/4522715:55
kgriffshmm16:03
kgriffshttp://arstechnica.com/information-technology/2013/09/guis-are-for-wimps-amazons-cloud-unveils-a-command-line/16:03
kgriffshasn't openstack had this for, like, ever?16:03
kgriffsmaybe not all services, I guess16:03
kgriffsoz_akan_: can you test this (just added a new commit): https://github.com/kgriffs/marconi16:04
kgriffsI need to know whether passing slave_ok to the find method is necessary16:05
kgriffsin the latest commit I removed it16:05
kgriffs(we know the patch fixes the bug when slave_ok=False already from your earlier test)16:05
openstackgerritFlavio Percoco proposed a change to stackforge/marconi: Implement small http client for tests  https://review.openstack.org/4526716:20
openstackgerritFlavio Percoco proposed a change to stackforge/marconi: Pull actual tests out of marconi/tests  https://review.openstack.org/4447516:20
openstackgerritFlavio Percoco proposed a change to stackforge/marconi: Implement embedded marconi-server execution  https://review.openstack.org/4474916:20
openstackgerritFlavio Percoco proposed a change to stackforge/marconi: Remove util packages  https://review.openstack.org/4447616:20
openstackgerritFlavio Percoco proposed a change to stackforge/marconi: Use oslo.config for functional tests  https://review.openstack.org/4447716:20
openstackgerritFlavio Percoco proposed a change to stackforge/marconi: Run functional tests under tox  https://review.openstack.org/4472316:20
flaper87kgriffs: ^ :P16:20
kgriffsholy patches, batman!16:21
flaper87kgriffs: LOOOOOOOOOOOOOL16:24
flaper87kgriffs: I'm pretty sure I'm wrong BUT, why is this being implemented in the storage side instead of the transport side? https://review.openstack.org/#/c/42933/516:24
flaper87I'd have expected default limits being enforced in the transport layer16:25
flaper87s/enforced/defined/16:25
kgriffsbefore this patch, default limits were set by default kwarg values in the storage driver methods16:26
flaper87in fact, the test for default limits was done for the wsgi transport not in storage16:26
flaper87kgriffs: yeah, I agree with making it configurable, I just had this feeling about it being done in the wrong place16:27
kgriffsthe test is for min/max, not defaults, right?16:27
kgriffsdefaults is what to do if the request doesn't specify a value for, say, limit16:27
kgriffseach transport could define that, I suppose, and then the storage drivers would make the relevant kwargs required (vs optional)16:28
kgriffsanyway, seems like that would be a separate patch if we wanted to change things up16:28
kgriffs(since this patch is just replacing foo=10 with foo=config.default_foo16:28
flaper87yeah, I agree with that, I wanted to get your thoughts about this16:29
kgriffsok16:29
flaper87paging limits is something the API defines16:29
kgriffsit's worth discussing16:29
flaper87so, I'd expect the configuration being set by the transport16:29
flaper87I mean, the default16:29
flaper87damn, what's wrong with me16:29
kgriffsheh16:29
kgriffsgotchya16:30
kgriffsif you'd like to file a bug or something, we can look at it16:30
flaper87yup, wiiiilll do!16:30
flaper87brb, dinner16:30
kgriffsttfn16:30
kgriffsgoing to lunch. bbl16:39
*** kgriffs is now known as kgriffs_afk16:40
*** kgriffs_afk is now known as kgriffs17:40
*** kgriffs is now known as kgriffs_afk17:50
*** kgriffs_afk is now known as kgriffs17:59
openstackgerritKurt Griffiths proposed a change to stackforge/marconi: fix(storage.mongodb): Race condition when creating a claim  https://review.openstack.org/4529918:32
*** tedross has quit IRC18:33
openstackgerritKurt Griffiths proposed a change to stackforge/marconi: feat(storage): configurable default paging size  https://review.openstack.org/4293318:45
*** tedross has joined #openstack-marconi18:48
openstackgerritKurt Griffiths proposed a change to stackforge/marconi: fix: find_one() running slow (MongoDB)  https://review.openstack.org/4456018:49
openstackgerritKurt Griffiths proposed a change to stackforge/marconi: fix: Requests get slower when queues have a lot of messages  https://review.openstack.org/4434018:49
*** tedross has quit IRC19:10
*** tedross has joined #openstack-marconi19:10
openstackgerritKurt Griffiths proposed a change to stackforge/marconi: WIP: MongoDB indexing perf  https://review.openstack.org/4530719:13
kgriffsflaper87: https://review.openstack.org/#/c/45299/19:16
kgriffsI will review the test patches as well19:16
openstackgerritKurt Griffiths proposed a change to stackforge/marconi: WIP: MongoDB indexing perf  https://review.openstack.org/4530719:22
openstackgerritA change was merged to stackforge/marconi: feat(storage): configurable default paging size  https://review.openstack.org/4293319:26
*** tedross has quit IRC19:27
*** tedross has joined #openstack-marconi19:40
*** tedross has quit IRC20:03
*** tedross has joined #openstack-marconi20:16
*** oz_akan_ has quit IRC21:04
*** oz_akan_ has joined #openstack-marconi21:05
flaper87kgriffs: ping21:06
*** oz_akan_ has quit IRC21:09
kgriffshey21:17
kgriffswassup, man?21:17
kgriffsflaper87: pong21:19
flaper87kgriffs: hey21:20
flaper87:D21:20
flaper87kgriffs: do you think we can get a very small sharded environment isntalled somewhere?21:21
flaper87to test mongo backend w/ both envs21:21
kgriffssharded or you mean just a single replica set?21:21
flaper87kgriffs: 1 shareded and 1 using replica sets21:22
flaper87like 1 vm, everything in there21:22
kgriffsoic21:22
flaper87I'm not thinking about testing performance but the driver itself21:22
kgriffsI don't see why not.21:22
flaper87like queries, configs, etc21:22
flaper87cool21:22
kgriffswe can host it at Rackspace or do you have something at RedHat in mind?21:23
kgriffss/RedHat/Red Hat21:23
kgriffss/Red Hat/RH21:23
kgriffs;p21:24
flaper87heheheh, erm, well, I was actually thinking at RAX since we've some tests VMS there already, but I guess I can figure something else out if needed21:25
kgriffsheh21:25
kgriffsno, I'm cool with it, just wondering if you had something slated for it already21:26
kgriffsLet me ask oz or malini to spin something up21:26
flaper87kgriffs: cool, I've nothing ready but I can definitely help setting it up21:26
kgriffsok, let me bring up an instance and then I'll give you the creds21:28
kgriffsI guess we had better just set it up to use pki instead of username/password and then we can add people's pub keys21:29
kgriffsflaper87: do you prefer CentOS or Ubuntu?21:30
* kgriffs chucles21:30
* kgriffs chuckles some more21:30
flaper87kgriffs: LOOOOOOL21:30
flaper87kgriffs: windows, pls21:30
kgriffsoh, sweet. I was hoping you'd say that. Windows ME should be perfect.21:34
*** tedross has quit IRC22:04
*** oz_akan_ has joined #openstack-marconi22:06
*** oz_akan_ has quit IRC22:26
*** amitgandhi has quit IRC22:27
openstackgerritKurt Griffiths proposed a change to stackforge/marconi: fix(transport.wsgi): JSON home doc missing GET and HEAD hints  https://review.openstack.org/4534122:40
*** kgriffs is now known as kgriffs_afk23:03
*** oz_akan_ has joined #openstack-marconi23:35
*** flaper87 is now known as flaper87|afk23:37
*** oz_akan_ has quit IRC23:40
*** oz_akan_ has joined #openstack-marconi23:40
*** oz_akan_ has quit IRC23:45
*** amitgandhi has joined #openstack-marconi23:54
« Wednesday, 2013-09-04 Index Friday, 2013-09-06 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!