Wednesday, 2014-02-05

« Tuesday, 2014-02-04 Index Thursday, 2014-02-06 »
*** jd__ has quit IRC04:21
*** jd__ has joined #openstack-climate04:22
*** Nikolay_St has joined #openstack-climate05:45
*** DinaBelova_ is now known as DinaBelova07:19
*** ppetit has joined #openstack-climate07:43
*** YorikSar has quit IRC08:10
*** YorikSar has joined #openstack-climate08:11
*** bauzas has joined #openstack-climate08:26
bauzasDinaBelova: I just noticed one bug with duplicate entries within my patch08:43
bauzasDinaBelova: it's V1 related, so could you please run on your own against trunk ?08:44
bauzasDinaBelova: when creating a DuplicateEntry for the lease name, it returns null with code 202 instead of an API Exception08:45
bauzasthat's pretty bad08:45
bauzasDinaBelova: and dunno why it hasn't been spotted by unittests08:45
DinaBelovabauzas, it might be, don't know relly08:46
DinaBelovaif so, please file bug in Launchpad08:46
bauzasI'm on my pecan branch08:46
bauzasthat's why I'm asking you to reproduce against master08:46
DinaBelovaOk, i'll check it08:46
bauzasfor some reason, it seems now that create_lease returns None if a duplicate entry08:47
bauzasinstead of raising a DB exception08:47
bauzasas if it was catched08:47
*** swann has joined #openstack-climate08:48
bauzasgot it08:48
bauzas...08:48
bauzashow come ?08:48
DinaBelova?08:48
DinaBelovacan't get your point08:49
bauzashttps://github.com/stackforge/climate/blob/master/climate/manager/service.py#L20408:49
bauzasswann: hi08:49
bauzasswann: seems like one of your patches produced a regression :D08:50
swannhi!08:50
bauzasDinaBelova: RuntimeError is now catched and only logged08:50
DinaBelovaI see that now08:50
swannbauzas: not at office today ?08:51
bauzasswann: wfh ;)08:51
bauzasdaughter at school to get back early08:51
bauzasswann: I was just noticing that DB errors are now catched on the Manager08:52
DinaBelovathat's a good thing to fix in 0.1.1 :)08:52
bauzasswann: consequently, when a DuplicateEntry comes, the API returns 202 with a lease equals to null08:52
DinaBelovaalso08:52
DinaBelovaI congratulate u folks with 0.1 :)08:52
DinaBelovaI'm writing release notes now08:53
bauzasDinaBelova: well, I think the good fix is to raise ClimateException instead of RuntimeError ;)08:53
DinaBelovaand will send maul to dev ml08:53
DinaBelovayep08:53
DinaBelovasure08:53
bauzashave you released the tarballs ?08:53
DinaBelovaon launchpad, yep08:53
bauzasgreat08:53
DinaBelovaI've uploaded them there08:54
swanncongrats to you deer ptl :)08:54
bauzasso you manually created the eggfiles ?08:54
DinaBelova:D08:54
DinaBelovadeer -> dear? :D:D:D08:54
bauzaspython setup.py build ?08:54
DinaBelova:D:D:D08:54
bauzass/build/bdist08:54
DinaBelovaI've used python setup.py sdist08:55
DinaBelovaand signed them with gpg key08:55
DinaBelovaas well as tags to repos08:55
bauzashow did you made the tags ?08:55
bauzasusing git ?08:56
DinaBelovapushed them with git, yep08:56
DinaBelovaas usual, although08:56
DinaBelova:)08:56
bauzasok got it08:56
bauzasso, tagged -> sdist -> git push ?08:57
bauzason github ?08:57
DinaBelova+ signing up08:57
DinaBelovaon gerrit08:57
bauzasok08:57
DinaBelovathta looks like git push gerrit tag08:57
bauzasof course08:57
bauzas:)08:57
DinaBelovathe only thing that is done automatically usually - uploading tarballs to pypi08:58
DinaBelovabut we have no way to do it now, so :)08:58
bauzas... ;)08:58
DinaBelovaalso I checked all tarballs that every needed file was there08:58
DinaBelovait turned out we had badly written setup.cfg's and no MANIFEST.in and LICENSE in every our repo08:59
DinaBelovaso i needed to fix that08:59
DinaBelovaafter I got good tarbolls08:59
DinaBelovaI've uploaded them08:59
bauzasgotcha08:59
bauzassaw the patches09:00
DinaBelova:)09:01
DinaBelovait was sudden thing :(09:01
DinaBelovai'm really sorry to give +2/A to myself :(09:02
bauzaswell, there was a matter of urgency09:04
DinaBelovasure09:04
bauzasfor other projects, only one +2/A is needed for emergency patches09:04
bauzasso for the next future, I'm OK with only Sergey to give +2/A in case of emergency09:05
DinaBelovathat's why i decided not to wait morning :D09:05
DinaBelovaSergey is not core :)09:05
bauzasoh, right09:05
DinaBelovaso he gave me +! :)09:05
DinaBelova+1, sorry09:05
DinaBelovaand checked me09:05
DinaBelova:)09:05
bauzaswell, we definitely need a US peer :)09:06
DinaBelova:D09:06
SergeyLukjanovI'm a east side US peer due to my activity time range ;)09:18
bauzasSergeyLukjanov: sleeping is not an option ;)09:35
SergeyLukjanovbauzas, exactly09:35
swannhey climaters, I'll work on https://bugs.launchpad.net/climate/+bug/1271875 but since the last ML discussion about gate failure due to external oslo lib config dependencies .. not sure we should use the generate/check_update sample config ...11:12
swannML thread : http://lists.openstack.org/pipermail/openstack-dev/2014-February/026213.html11:12
swannso, what we'll do in climate ?11:12
swannI propose to generate config sample but no use the check_update script in tests11:13
swannthoughts ?11:13
bauzasswann: well, the PEP8 check fails because of a gating test11:26
bauzashttps://github.com/openstack/nova/blob/master/tox.ini#L2711:26
swannbauzas: yes that's what I tried to tell ...  My question was do we need this test (in tox pep8 env) in Climate ?11:33
swannsubsequential question .. do we need to  generate our sample config with this tools .. ?11:34
swannanyway, I've the patch .. I push it and let talk team on the review11:34
swannbauzas: did you read the intersting ML thread about asyncio .. I guess you did :)11:36
*** ppetit has left #openstack-climate11:37
openstackgerritSwann Croiset proposed a change to stackforge/climate: Use config generator/updater tools from Oslo incubator  https://review.openstack.org/7124311:38
bauzasswann: I think we should put the run into a non-voting gate12:29
bauzasswann: and yes, I talked about asyncio yesterday, see my mail about Py312:30
bauzasI was at the asyncio FOSDEM presentation on Sunday12:30
*** casanch1 has joined #openstack-climate12:30
bauzasswann: maybe you were still assuming the yesterday's events :D12:31
*** ppetit has joined #openstack-climate12:41
swannbauzas: I was just behind you, don't remember ?12:51
swannwith big sunglasses12:52
DinaBelova:D13:03
DinaBelovayou're funny, guys :)13:03
bauzasswann: naaaaah ?13:04
DinaBelovaalso I'm not sure I know about asyncio13:04
DinaBelovamay you provide me some info?13:04
bauzasDinaBelova: well, what happens in Brussels stays in Brussels :D13:04
DinaBelova:D:D:D13:04
bauzasDinaBelova: asyncio is quite good for green threads because it provides new Future and Tasks13:05
bauzaswhich can be led asynchroniously13:05
DinaBelovani-i-ice!13:05
bauzasand the Python GIL has been improved13:05
bauzasso, that's a sexy thing13:05
DinaBelova=^_^=13:06
bauzasI sent an email to the -dev asking for py3 compat13:06
bauzasthere is basically an on-going stuff for oslo13:06
bauzasthe main blocking point is evenlet13:07
bauzaseventlet13:07
DinaBelovayep, I saw that....13:07
bauzasbut there is room for porting eventlet with undergoing asyncio13:07
DinaBelovaok, gotcha13:07
bauzasI'm not fan of trollius13:07
bauzaswhich is a backport of asyncio to py2.6 and above13:08
bauzasbecause there is only one guy supporting it13:08
bauzasfrom enovance13:08
DinaBelovabus factor is really low, yep13:08
bauzascompared to asyncio which will be stdlib in py3.413:08
DinaBelovathat will be much better13:09
DinaBelovabut 3.4...13:09
DinaBelovadunno really13:09
DinaBelovawe should look where this discussion will go13:09
bauzasi don't think that will impact us13:10
bauzasexcept that we will have to point to a latter eventlet13:11
DinaBelova+113:11
openstackgerritPablo Andres Fuente proposed a change to stackforge/python-climateclient: Fix PEP8 H302 import only modules  https://review.openstack.org/7128415:15
*** casanch1 has quit IRC15:21
*** casanch1 has joined #openstack-climate15:30
*** DinaBelova is now known as DinaBelova_15:53
swannDinaBelova: could you rebase the oslo.messaging patch to master please. Will be easier to test it.16:07
swannmany thanks16:07
openstackgerritPablo Andres Fuente proposed a change to stackforge/python-climateclient: Fix PEP8 H302 import only modules  https://review.openstack.org/7128416:11
*** pafuent has joined #openstack-climate16:13
*** DinaBelova_ is now known as DinaBelova16:26
*** bauzas has quit IRC16:40
pafuentSergeyLukjanov: Are you around?16:41
pafuentHi17:14
*** casanch1 has quit IRC17:15
*** casanch1 has joined #openstack-climate17:15
pafuentI'm doing a climate lease-list and sometimes I got a ProgrammingError complaining about that the leases table don't have the column trust17:18
pafuentI checked the models and I can't find that column17:19
pafuentAnd this happens every 3 times that I do lease-list17:24
DinaBelovapafuent, hello17:38
pafuentDinaBelova: Hi17:38
DinaBelovau might have old table for Climate....17:38
DinaBelovacause now we have no trust, we have trust_id field there17:39
DinaBelovaalso check please if you have more than one climate-manager running17:39
DinaBelovasuch strange errors might happen now too because of multiple managers servers17:40
pafuentThe models.py have the trust_id and leases table contains trust_id17:40
pafuentI'll check the managers thing17:40
DinaBelovaok17:40
pafuentI've at least five17:41
DinaBelova0_017:41
DinaBelovaoh17:41
DinaBelovatoo much :D17:41
DinaBelovakill them all and start one only :)17:41
DinaBelovaand try it one more time :)17:41
DinaBelovaare you using climate with some kind of devstack? cause now we have no fully working devstack version17:42
DinaBelovaonly drafts17:42
pafuentNow works17:42
pafuentThanks17:42
DinaBelovaso climate services might not be killed while unstashing17:42
DinaBelovau're welcome :)17:42
pafuentYeap17:42
pafuentBTW, At which time I can found SergeyLukjanov at IRC?17:43
DinaBelovaHe's almost everyday here :) I suppose he'll be back in an hour17:44
DinaBelovaor even quicker17:44
pafuentOk, thanks so much17:44
DinaBelovahe's like me in Moscow time zone17:44
DinaBelovabut he's online almost every day till midnight (that's 20:00 UTC)17:45
DinaBelovapafuent, although I suppose that I can try to help you with Climate speaking about all its problems :D17:46
DinaBelovaas Climate PTL I know something about it ;)17:46
pafuentDinaBelova: ROFL17:47
pafuentDinaBelova: It's related to his patch for climate on devstack17:47
DinaBelovayep, I got it17:47
DinaBelovaI suppose I'll continue working on it17:48
pafuentpafuent: I added some comments to that review and then I notice that seems abandoned17:48
DinaBelovaSergey is too busy 'cause of Savanna's related changes17:48
DinaBelovaI'll restore it tomorrow and will continue working on it17:48
DinaBelovawe hoped he'll finish this change himself, but he has no time now for that17:49
pafuentpafuent: And I want to ask him if I could work on that patch17:49
DinaBelovaoh :)17:49
DinaBelovaypu'd like to?17:49
DinaBelovait will be really nice for you :)17:49
DinaBelovaI'm ok with that - I'll ask him to restore it, and I believe you can finidh it17:50
pafuentDinaBelova: Yep. I don't know a lot about doing that kind of stuff but would like to test myself against that17:50
DinaBelovafinish, sorry17:50
DinaBelovaI guess it's a good idea :)17:50
pafuentpafuent: Is still open, so I could download the patch with gerrit and amend it, right?17:51
DinaBelovasmth like that, yes17:52
DinaBelovait's wip now17:52
DinaBelovaso you can just get it and amned17:53
DinaBelovaamend, sorry17:53
DinaBelovathe only moment here is that it'll be SL as author there17:53
DinaBelovai guess it's good to use co-author there17:53
DinaBelovaSergey will tell you how (he'll be here in several mins)17:54
pafuentDinaBelova: Ok, thanks17:54
DinaBelova:)17:54
SergeyLukjanovpafuent, DinaBelova, hey18:01
SergeyLukjanovI'm sorry about my devstack patch18:01
SergeyLukjanovit was a first iteration :)18:01
SergeyLukjanovpafuent, you can just take my diff and create own CR18:02
SergeyLukjanovor just write it from scratch18:02
DinaBelovaI suppose Pablo may just continue working on yours18:02
pafuentI think it's better to continue your work18:03
pafuentHow I should handle the co-author thing?18:03
SergeyLukjanovpafuent, example: https://github.com/openstack/savanna/commit/07d033dfcbe497c78e651b9085e5712eff743a5518:04
pafuentSergeyLukjanov: Ok, so I should put my name/email using the Co-Authored-By in the commit message18:06
SergeyLukjanovpafuent, exactly18:07
SergeyLukjanovI've added few more notes to this patch18:07
DinaBelova:D18:08
pafuentSergeyLukjanov: Ok, I'll check them18:08
DinaBelovaSergeyLukjanov, 'if it's not mine anymore let's post some more comments!' :)18:08
DinaBelovahahaha18:08
SergeyLukjanovDinaBelova, heh :)18:09
*** casanch1 has quit IRC18:12
openstackgerritPablo Andres Fuente proposed a change to stackforge/python-climateclient: Support not-json error responses  https://review.openstack.org/7132818:14
*** casanch1 has joined #openstack-climate18:15
*** bauzas has joined #openstack-climate18:48
pafuentSergeyLukjanov: Do you know if create the keystone v3 endpoint during the climate devstack setup it's fine?19:07
SergeyLukjanovpafuent, it shouldn't break anythong19:08
SergeyLukjanovpafuent, and while it'll be used only by climate it's ok19:08
SergeyLukjanovIMO the better approach is to take identity service URL and replace v2.0 with v3 in Climate19:09
*** f_rossigneux has quit IRC19:10
pafuentSergeyLukjanov: Ok, Thanks19:10
*** f_rossigneux has joined #openstack-climate19:49
DinaBelovaSergeyLukjanov, that will be done soon in Climate19:52
DinaBelovawhen it will be we'll have no need in v3 endpoint19:52
*** casanch1 has quit IRC20:18
*** pafuent has left #openstack-climate21:03
*** DinaBelova is now known as DinaBelova_21:09
*** bauzas has quit IRC22:17
« Tuesday, 2014-02-04 Index Thursday, 2014-02-06 »

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