Friday, 2015-04-24

« Thursday, 2015-04-23 Index Saturday, 2015-04-25 »
*** sdake has quit IRC00:03
*** tsekiyam_ has joined #openstack-oslo00:06
*** salv-orlando has quit IRC00:07
*** achanda has quit IRC00:07
*** mtanino has quit IRC00:08
*** salv-orlando has joined #openstack-oslo00:08
*** tsekiyama has quit IRC00:09
*** tsekiyam_ has quit IRC00:11
*** zzzeek has quit IRC00:11
*** vilobhmm11 has quit IRC00:15
*** vilobhmm1 has joined #openstack-oslo00:15
*** salv-orlando has quit IRC00:22
openstackgerritJoshua Harlow proposed openstack/taskflow: Start building a redis jobboard (WIP)  https://review.openstack.org/17695100:28
openstackgerritMerged openstack-dev/pbr: Be safe, don't let friends use RawConfigParser  https://review.openstack.org/17699700:53
openstackgerritJoshua Harlow proposed openstack/taskflow: Start building a redis jobboard (WIP)  https://review.openstack.org/17695100:54
*** haigang has joined #openstack-oslo01:15
*** bknudson has joined #openstack-oslo01:21
*** ChuckC has quit IRC01:21
*** sdake has joined #openstack-oslo01:22
*** sdake_ has quit IRC01:26
*** flaper87 has quit IRC01:27
*** flaper87 has joined #openstack-oslo01:28
*** liusheng has joined #openstack-oslo01:29
*** flaper87 has quit IRC01:33
*** salv-orlando has joined #openstack-oslo01:35
*** flaper87 has joined #openstack-oslo01:39
*** vilobhmm1 has quit IRC01:40
*** vilobhmm1 has joined #openstack-oslo01:40
*** vilobhmm1 has quit IRC01:40
*** salv-orlando has quit IRC01:45
*** ChuckC has joined #openstack-oslo01:49
*** sdake_ has joined #openstack-oslo01:50
*** shardy_ has joined #openstack-oslo01:50
*** shardy has quit IRC01:51
*** sdake has quit IRC01:53
*** ChuckC has quit IRC01:53
*** ChuckC has joined #openstack-oslo01:54
*** Kennan2 is now known as Kennan01:55
*** shardy_ has quit IRC01:55
*** shardy has joined #openstack-oslo01:55
*** harlowja is now known as harlowja_away02:02
*** kbyrne has quit IRC02:05
*** kbyrne has joined #openstack-oslo02:10
*** salv-orlando has joined #openstack-oslo02:15
*** salv-orlando has quit IRC02:20
*** achanda has joined #openstack-oslo02:37
*** haigang has quit IRC02:58
*** sdake has joined #openstack-oslo03:00
*** sdake_ has quit IRC03:04
*** sdake_ has joined #openstack-oslo03:04
*** sdake has quit IRC03:08
*** yamahata has joined #openstack-oslo03:23
*** liusheng has left #openstack-oslo03:24
*** rushiagr_away is now known as rushiagr03:40
*** achanda has quit IRC03:44
*** vilobhmm1 has joined #openstack-oslo03:57
*** rushiagr is now known as rushiagr_away04:00
*** rushiagr_away is now known as rushiagr04:02
openstackgerritJoshua Harlow proposed openstack/taskflow: Expose fake filesystem 'join' and 'normpath'  https://review.openstack.org/17651504:15
*** salv-orlando has joined #openstack-oslo04:18
*** salv-orlando has quit IRC04:25
openstackgerritJoshua Harlow proposed openstack/taskflow: Start building a redis jobboard (WIP)  https://review.openstack.org/17695104:27
*** subscope_ has joined #openstack-oslo04:30
*** achanda has joined #openstack-oslo04:35
*** takedakn has joined #openstack-oslo04:45
*** takedakn has quit IRC04:45
*** nkrinner has joined #openstack-oslo04:52
*** salv-orlando has joined #openstack-oslo04:58
*** browne has joined #openstack-oslo05:03
openstackgerritMerged openstack/taskflow: Add + use diagram explaining retry controller area of influence  https://review.openstack.org/17649605:04
openstackgerritMerged openstack/taskflow: Add openclipart.org conductor image to conductor docs  https://review.openstack.org/17655905:05
*** salv-orlando has quit IRC05:05
*** rushiagr is now known as rushiagr_away05:06
*** amrith is now known as _amrith_05:11
*** salv-orlando has joined #openstack-oslo05:17
openstackgerritMerged openstack/taskflow: Test more engine types in argument passing unit test  https://review.openstack.org/17650205:21
openstackgerritMerged openstack/taskflow: Replace more instance(s) of exception chaining with helper  https://review.openstack.org/17559305:21
openstackgerritMerged openstack/taskflow: Use oslo_utils eventletutils to warn about eventlet patching  https://review.openstack.org/17655405:21
*** _amrith_ is now known as amrith05:21
*** salv-orlando has quit IRC05:21
*** inc0 has joined #openstack-oslo05:26
*** arnaud___ has joined #openstack-oslo05:39
*** e0ne has joined #openstack-oslo05:43
*** e0ne is now known as e0ne_05:43
*** e0ne_ is now known as e0ne05:43
*** vilobhmm1 has quit IRC05:44
*** e0ne has quit IRC05:45
*** takedakn has joined #openstack-oslo05:53
*** rushiagr_away is now known as rushiagr06:01
*** david-ly_ has joined #openstack-oslo06:01
*** david-lyle has quit IRC06:03
*** subscope_ has quit IRC06:11
openstackgerritOpenStack Proposal Bot proposed openstack/oslo.utils: Imported Translations from Transifex  https://review.openstack.org/17663906:13
*** stevemar has quit IRC06:32
*** arnaud___ has quit IRC06:33
openstackgerritMerged openstack/taskflow: Expose fake filesystem 'join' and 'normpath'  https://review.openstack.org/17651506:35
*** takedakn has quit IRC06:42
*** salv-orlando has joined #openstack-oslo06:45
*** salv-orlando has quit IRC06:55
*** haypo has joined #openstack-oslo07:09
*** salv-orlando has joined #openstack-oslo07:12
*** takedakn has joined #openstack-oslo07:12
*** salv-orlando has quit IRC07:14
*** sdake_ has quit IRC07:20
*** achanda has quit IRC07:24
*** achanda has joined #openstack-oslo07:36
*** salv-orlando has joined #openstack-oslo07:40
*** browne has quit IRC07:42
*** e0ne has joined #openstack-oslo07:47
*** e0ne has quit IRC07:49
*** e0ne has joined #openstack-oslo07:53
openstackgerritKamil Rykowski proposed openstack/oslo.middleware: Update response body when healthcheck is available  https://review.openstack.org/17711007:54
*** achanda has quit IRC07:57
*** e0ne has quit IRC08:05
*** ajo has joined #openstack-oslo08:06
*** e0ne has joined #openstack-oslo08:09
*** ajo_ has joined #openstack-oslo08:09
*** ajo has quit IRC08:09
openstackgerritKamil Rykowski proposed openstack/oslo.middleware: Update response body when healthcheck is available  https://review.openstack.org/17711008:10
*** e0ne has quit IRC08:11
*** yassine_ has joined #openstack-oslo08:21
*** ihrachyshka has joined #openstack-oslo08:23
*** andreykurilin__ has joined #openstack-oslo08:27
*** ndipanov has joined #openstack-oslo08:37
*** ndipanov is now known as ndipanoff08:38
openstackgerritVictor Sergeyev proposed openstack/oslo.messaging: Add JsonPayloadSerializer serializer  https://review.openstack.org/17689308:39
*** ozamiatin has joined #openstack-oslo08:40
*** salv-orlando has quit IRC08:43
*** ozamiatin has quit IRC08:50
*** ajo_ has quit IRC08:51
*** ajo has joined #openstack-oslo08:59
*** inc0_ has joined #openstack-oslo09:02
*** ndipanoff has quit IRC09:02
openstackgerritOpenStack Proposal Bot proposed openstack/oslo.messaging: Updated from global requirements  https://review.openstack.org/17712909:02
*** ozamiatin has joined #openstack-oslo09:05
*** inc0 has quit IRC09:06
*** e0ne has joined #openstack-oslo09:10
*** ajo has quit IRC09:10
*** ndipanov has joined #openstack-oslo09:10
ndipanovmdbooth, sorry my laptop died since I forgot to plug it in09:11
ndipanovmdbooth, fwiw - we backported the test_tracker.py to RHOS09:11
mdboothndipanov: I forgive you09:11
ndipanovso if you are actually doing this to backport it to rhos 609:11
ndipanovall the hard work of backporting the tests upstream is useless on rhos for that particular case09:12
ndipanovbut for filters it's not09:12
*** ndipanov is now known as ndipanoff09:13
*** e0ne is now known as e0ne_09:16
*** ajo has joined #openstack-oslo09:16
*** e0ne_ has quit IRC09:21
*** ozamiatin has joined #openstack-oslo09:28
*** e0ne has joined #openstack-oslo09:35
*** salv-orlando has joined #openstack-oslo09:35
openstackgerritIhar Hrachyshka proposed openstack-dev/hacking: Add support for flake8 off_by_default for optional checks  https://review.openstack.org/13405209:36
*** sdake has joined #openstack-oslo09:37
*** sdake_ has joined #openstack-oslo09:39
*** sdake has quit IRC09:43
*** jaosorior has joined #openstack-oslo09:46
*** ozamiatin has quit IRC09:47
*** cdent has joined #openstack-oslo09:57
*** flaper87 has quit IRC10:05
*** flaper87 has joined #openstack-oslo10:05
*** inc0_ has quit IRC10:18
*** inc0 has joined #openstack-oslo10:21
*** andreykurilin__ has quit IRC10:26
*** ozamiatin has joined #openstack-oslo10:26
*** inc0 has quit IRC10:31
*** sdake_ has quit IRC10:53
*** e0ne is now known as e0ne_11:13
*** cdent has quit IRC11:20
*** alexpilotti has joined #openstack-oslo11:20
*** e0ne_ has quit IRC11:23
*** dguitarbite has quit IRC11:24
*** eezhova has quit IRC11:24
*** dguitarbite has joined #openstack-oslo11:25
*** inc0 has joined #openstack-oslo11:28
*** yamahata has quit IRC11:31
*** inc0 has quit IRC11:33
*** inc0_ has joined #openstack-oslo11:33
*** ozamiatin has quit IRC11:35
openstackgerritMerged openstack-dev/hacking: Updated from global requirements  https://review.openstack.org/17682011:40
haypodhellmann, jd__ : oh, requirements is not more frozen in master? if yes, could you please review my eventlet upgrade for python3? https://review.openstack.org/#/c/172132/ it should unblock two pending python 3 patches for oslo messaging11:43
*** e0ne has joined #openstack-oslo11:43
*** inc0_ has quit IRC11:44
haypowith my 3 pending oslo.messaging patches, we will having oslo.messaging working on python 3 for all executors and drivers! ... except of Qpid and AMQP 1.0 drivers11:44
haypo(any volunteer to port these drivers? :-))11:45
*** inc0 has joined #openstack-oslo11:45
*** inc0 has quit IRC11:51
*** dguitarbite has quit IRC11:53
*** EmilienM has quit IRC12:01
*** EmilienM has joined #openstack-oslo12:01
*** exploreshaifali has joined #openstack-oslo12:04
*** kgiusti has joined #openstack-oslo12:09
*** subscope has quit IRC12:11
*** tedross has joined #openstack-oslo12:13
*** tedross has left #openstack-oslo12:19
hayposileht, jd__ : my  remaining oslo.messaging patches: https://review.openstack.org/#/c/172135/ (Enable eventlet dependency on Python 3, it was approved, but i had to rebase it) & https://review.openstack.org/#/c/175759/ (test_matchmaker_redis)12:23
*** subscope has joined #openstack-oslo12:25
openstackgerritVictor Stinner proposed openstack/oslo.messaging: Enable eventlet dependency on Python 3  https://review.openstack.org/17213512:29
haypooh, the changelog was outdated, it's now fixed ;-) ^^12:29
haypoI wrote a long comment to justify the futures requirements on Python 3 (workaround a bug in tox). i have no solution for it right now :-/12:29
haypotox developers don't want to fix, and environment markers are not supported yet...12:30
hayposileht, oops, you reviewed patch set 6, but i posted a new version because the changelog was outdated (it described an old version of the patch set): https://review.openstack.org/#/c/172135/12:31
hayposileht, are you aware of anyone interested to port QPID/AMQP to python3?12:32
silehthaypo, no12:32
hayposileht, too bad :-p12:32
silehthaypo, for AMQP1 once the proton c lib will be generated to python3, that should be easy12:33
hayposileht, there is some C code which need to be ported to Python 3, https://git-wip-us.apache.org/repos/asf?p=qpid-proton.git;a=blob;f=proton-c/bindings/python/cproton.i;h=70d6c42fa5e0b67ef0a9f85b2361c15bc9607990;hb=HEAD12:33
hayposileht, it's easy, we just have to find a volunteer ;)12:34
silehtas usual...12:34
hayposileht, i ported eventlet, paste, python-memcached this week. i may port qpid/amqp later :-p12:35
silehthaypo, only rabbit is fully tested/supported12:35
hayposileht, i read recently that qpid is not tested by gates, is that right?12:36
silehthaypo, I think for oslo.messaging, we are pretty good now for python3 stuffs12:36
silehthaypo, it only have functionnal testing12:36
hayposileht, last days, i repeated multiple that "it's almost done", but each time i found new issues... (like python-memcached, i forgot this one)12:36
*** dguitarbite has joined #openstack-oslo12:36
hayposileht, zigo mentionned pysaml2 for keystone which doesn't support py312:37
zigohaypo: Yeah, and I'm working on it right now.12:37
hayposileht, ah functional tests are running with qpid? cool12:38
haypozigo, hey thomas!12:38
zigohaypo: There's a bunch of Except missing the as thing, lots of print statements, and using urlparse instead of urllib.parse.12:38
haypozigo, i didn't know that you were on this channel :-D12:38
zigohaypo: I'm everywhere! :)12:38
zigohaypo: Are you comming to Vancouver?12:38
haypozigo, try 2to6 ;)  https://github.com/haypo/2to6/ (it's not my project, i made minor changes)12:38
haypozigo, yes, i do12:39
haypozigo, to promote python3 ;)12:39
zigoSuper cool!12:39
zigohaypo: Paultag created a python3 porter team in Debian.12:39
zigohaypo: The goal is to completely remove Python2 from Debian Stretch (FYI: Jessie, Buster, Stretch, in this order).12:39
haypozigo, one of your colleague contacted to help me on porting nova to python 3 ;) Eugeniya Kudryashova12:39
zigoOh ok.12:40
haypozigo, now i have to convince my own colleagues that python3 rocks ;)12:40
zigohaypo: We are about 1000 in Mirantis, so I don't know everyone! :)12:40
haypozigo, you must know all your colleagues, as i know all 7000 redhatters12:40
haypowell right now, i only know them by their employee number, not their name yet :-D12:41
*** bknudson has quit IRC12:42
zigo:)12:46
zigohaypo: Will you work on python-qpid too?12:46
zigoThat's a blocker for oslo.messaging ...12:46
zigoI personally would have prefer if we just removed qpid.12:47
zigoBut that's not my call ...12:47
haypozigo, "I personally would have prefer if we just removed qpid." ask  sileht  for that :)12:47
zigoWhy that? He's involved in qpid? :)12:48
haypozigo, "That's a blocker for oslo.messaging ..." i don't think so. eventlet, qpid, etc. are optional dependencies12:48
haypozigo, involved in olso messaging12:48
zigohaypo: There's unit tests with it.12:48
zigohaypo: Or are you suggesting that I just blacklist the qpid unit tests, and be done with it? :)12:48
zigoThat's of course one possibility.12:48
haypozigo, i didn't suggest anything12:49
haypozigo, oh, you mean for the packaging? is qpid a requirement for your oslo.messaging package?12:49
haypozigo, it's common to skip tests if a module is missing12:49
haypozigo, python-keystoneclient skips memcache tests if memcache is missing, and memcache is not in requirements, so...12:50
zigohaypo: It's as a build-depends, yes, because it's used by unit tests.12:50
haypozigo, skip these tests on python312:50
zigohaypo: Not the case in oslo.messaging, I'd have to manually blacklist it, and I believe it would be a problem for the subunit list if qpid isn't there.12:50
zigohaypo: Just skipping tests in python3 wouldn't work, as we would still need the module for listing tests.12:51
zigo(so subunit --list would crash in python 3)12:51
dstanekit looks like Keystone will be moving away from python-memcache in the future, but i'm not sure exactly when12:51
zigodstanek: That's not what morganfainberg wrote. He'd be ok with that, but it's far from his priority. His priority is having the middleware moved away from keystoneclient.12:52
haypodstanek, because it doesn't support python3, but i read that pymemcache is better12:52
zigohaypo: In python3, I get: ImportError: No module named 'time_util'12:53
haypo(ah yes, and the middleware should be removed from the client)12:53
zigoHas this been removed too?12:53
haypozigo, in which project?12:53
zigopysaml212:53
*** e0ne is now known as e0ne_12:53
dstanekzigo: last i heard it was definitely a move we were going to make, but there was no timeline because it's low priority12:54
zigohaypo: Ah, shit, in pysaml2, url.parse doesn't work in Python 2 ... :(12:54
dstanekworking on getting Keystone working on Python3 as we speak12:54
zigohaypo: Do I have to do something like try: import urlparse except: import urllib.parse?12:54
*** e0ne_ is now known as e0ne12:55
haypozigo, use six.moves for that12:55
zigohaypo: How?12:55
haypozigo, don't hesitate to complete https://wiki.openstack.org/wiki/Python3#Port_Python_2_code_to_Python_3 ;-)12:55
haypozigo, pick the right module in https://pythonhosted.org/six/#module-six.moves12:55
*** yassine_ has quit IRC12:56
haypozigo, for example, "from six.moves import urllib" and then use "urllib.parse.(...)" or "urllib.request.(...)"12:56
openstackgerritOpenStack Proposal Bot proposed openstack-dev/hacking: Updated from global requirements  https://review.openstack.org/17722912:56
zigohaypo: Thanks!12:56
dstanekdoes oslo-messaging now work with Py3?12:58
haypodstanek, short answer: no12:58
haypodstanek, i will in a few days12:58
*** yassine has joined #openstack-oslo12:59
dstanekhaypo: ok, thx. i'll keep that mocked out for now12:59
haypodstanek, put your +1 on https://review.openstack.org/#/c/175759/ & https://review.openstack.org/#/c/172135/ ;)12:59
dstanekhaypo: ok, i'll take a look in a few12:59
*** nkrinner has quit IRC13:03
ihrachyshkadhellmann, I have some oslo patches needed for neutron that would need some love. I won't be able to raise them on Mon the next two weeks since I'll be on vacation, so I've updated agenda: https://wiki.openstack.org/wiki/Meetings/Oslo#Agenda_for_Next_Meeting13:03
openstackgerritOpenStack Proposal Bot proposed openstack/oslo.vmware: Updated from global requirements  https://review.openstack.org/17723913:04
openstackgerritOpenStack Proposal Bot proposed openstack/oslo.versionedobjects: Updated from global requirements  https://review.openstack.org/17724013:04
*** gordc has joined #openstack-oslo13:04
*** ajo has quit IRC13:05
zigohaypo: What's the point of adding trollius and aioeventlet if eventlet 0.17.3 is supposed to be Py3 compatible?13:11
*** ajo has joined #openstack-oslo13:14
haypozigo, come on...13:14
*** stpierre has joined #openstack-oslo13:14
haypozigo, i'm tired to explaining the same things again and again13:14
zigoSorry ...13:14
* zigo hides13:14
haypozigo, replacing eventlet with asyncio and porting openstack to python 3 are two separated projects13:15
haypozigo, it's not possible to use asyncio because it requires python 3.3+13:15
zigohaypo: I was asking, in the context of this patch which is in review.13:15
zigohaypo: I personally don't care much about Python 3.3, Jessie is out TOMORROW ! :)13:16
haypozigo, hum, maybe you question why: do we still need to replace eventlet? => the rationale is in https://review.openstack.org/#/c/164035/ (yes, eventlet must die)13:16
haypozigo, i mean that asyncio cannot be use in openstack because except of gnocchi, no openstack application works on python 313:16
zigoYes, eventlet must die, I agree.13:17
zigohaypo: Something not well known enough: with eventlet, we have an SSL downgrade attack hole on the paste daemons of OpenStack...13:18
*** ozamiatin has joined #openstack-oslo13:19
zigohaypo: The fun thing about porting to Py3, is that I may do correctly it without even knowing what the stupid code is doing ! :)13:20
haypozigo, you're right. i'm not aware of that. why is it specific to eventlet?13:20
haypozigo, haha, yes. i ported such much python 3 last weeks. i don't know anything about them13:20
haypozigo, ian bicking made me the new Paste maintainer whereas i never used this project...13:21
zigohaypo: If you use it for clients (which is the case for many clients), then it's going to use the SSL23 object of Python, even if you specify TLS.13:21
zigohaypo: Because it thinks it "knows better" and believe that it's ok to use any protocol which the server will agree on.13:21
zigoThen badaboum, poodle ...13:21
zigohaypo: Do a grep on "SSL23" on eventlet, and you'll see !13:22
dstanekzigo: ssl terminator ftw!13:22
zigodstanek: Exactly, but that's orthogonal to the issue.13:22
zigodstanek: What if you want to use SSL in your LAN where OpenStack is setup?13:22
zigo(ie: between projects and keystone...)13:22
dstanekzigo: use a terminator13:23
*** zz_jgrimm is now known as jgrimm13:23
zigoSo yes, we should HAProxy to do the work for clients to connect, but that's not it...13:23
dstaneki would use something like stud or stunnel13:23
zigodstanek: The reality is that everyone will *not* use SSL, and do management traffic over http only.13:24
dstanekzigo: they have that right, but if they want to do it correctly they'd run a terminator in front of each service13:24
*** jgrimm is now known as zz_jgrimm13:25
zigodstanek: The point is, we shouldn't provide something THAT broken, and it's completely false to just expect everyone to "know better".13:25
zigoSo, let's kill eventlet even more ! :)13:25
dstaneki don't disagree13:26
dstanekkeystone recommends deploying behind apache13:26
haypozigo, SSLv23 is only the default value13:28
zigohaypo: I don't think so.13:28
haypozigo, python uses the same default value13:28
zigo(but I'd have to check again...)13:28
haypozigo, https://github.com/eventlet/eventlet/blob/master/eventlet/green/ssl.py#L4913:29
zigohaypo: Look at the end of eventlet/convenience.py and see...13:31
zigocontext = SSL.Context(SSL.SSLv23_METHOD)13:31
zigoNo variable involved ...13:31
haypozigo, "convenience.py" doesn't sound like a mandatory module :)13:32
zigohaypo: But that's were we ended killing a unit test of Keystone because of it.13:33
zigo(that unit test was about checking SSL cert...)13:33
*** mriedem_away is now known as mriedem13:35
*** jungleboyj has quit IRC13:36
*** bknudson has joined #openstack-oslo13:44
*** rushiagr is now known as rushiagr_away13:44
dstanekhaypo: did you say you were already working on pysaml2?13:48
*** mriedem has quit IRC13:56
*** stevemar has joined #openstack-oslo13:57
zigohaypo: What with the time_util module in Python 3?13:59
*** dansmith is now known as superdan13:59
zigoHas it been renamed or something?13:59
zigodstanek: I am.14:00
haypodstanek, "<dstanek> haypo: did you say you were already working on pysaml2?" nope, i didn't touch this project14:01
*** mriedem has joined #openstack-oslo14:01
dstanekzigo: cool. i'm running into some errors getting keystone tests running on py3 and that lib is why :-)14:02
zigodstanek: Which is what's pushing me to work on it.14:02
*** sigmavirus24_awa is now known as sigmavirus2414:03
dstaneki think i have most of the Keystone code fixed - now just waiting on the deps14:03
haypozigo, i don't know time_util. where does it come from?14:04
haypozigo, do you write "import time_util", or does it come from a library?14:04
dstaneki think it's from pysaml214:04
zigohaypo: dstanek is right.14:05
zigoFrom pysaml214:05
haypozigo, hum, so what is your question?14:05
dstanekpysam2 likely isn't doing absolute imports - so it looks like it doesn't exist in python314:05
zigohaypo: Here's my error in Py3:14:05
*** takedakn has quit IRC14:05
zigosrc/saml2/validate.py:5: in <module>14:05
zigo    import time_util14:05
zigoE   ImportError: No module named 'time_util'14:05
zigoHow do I fix that?14:05
*** ChuckC_ has joined #openstack-oslo14:06
dstanekzigo: change the import to 'from saml2 import time_util'14:06
haypozigo, oh. it's a common trap of relative imports14:06
zigoAh...14:06
haypozigo, use "from . import time_util" or "from pysaml2 import time_util"14:06
hayposomething like that14:06
zigoThanks.14:06
*** ChuckC has quit IRC14:06
dstaneki've been adding 'from __future__ import absolute_import' to the top of most of the Keystone modules to catch things like that14:06
haypodstanek, it helps, but it's annoying to add it everywhere :)14:07
dstanekhaypo: yes, but good practice while supporting 2 and 3 in the same code base14:08
haypodstanek, ok14:08
dstanekfor something as small as pysaml2 i think modernize would fix all of the issues14:09
dstanekit'll also find things you won't find by hand like py2 .keys() vs. py3 .keys()14:09
haypodstanek, try also 2to6! https://github.com/haypo/2to6/14:11
haypodstanek, it should catch most obvious issues14:12
haypoand modify directly the code to use six14:12
dstanekthat's exactly what modernize does14:12
haypodstanek, oh, so modernize probably does the same changes than 2to614:12
*** ChuckC_ has quit IRC14:13
openstackgerritMerged openstack/oslo.messaging: Fix test_matchmaker_redis on Python 3  https://review.openstack.org/17575914:26
*** mtanino has joined #openstack-oslo14:27
*** ChuckC_ has joined #openstack-oslo14:28
haypowhen gate-tempest-dsvm-largeops-src-oslo.messaging fails 2 times, is it acceptable to ask again reverify?14:28
zigoHow can someone be so stupid as to store a file_old.py when using Git? :(14:29
haypozigo, yeah, rename it to file_old.py.bak14:29
zigo:)14:29
*** dguitarbite has quit IRC14:30
zigohaypo: What's the best way to print in stderr that would be compatible with python2 and 3?14:31
zigofile=sys.stderr doesn't work in python2, right?14:31
dstanekzigo: it should14:31
haypozigo, from __future__ import print_function14:32
haypozigo, then use print(..., file=sys.stderr)14:32
haypozigo, it works on python 2.6+14:32
zigoOh ok.14:32
haypozigo, be careful, in one hour, you will be an expert to port code to py314:32
dstanekah, doh. i didn't read the line about print before it14:32
*** zz_jgrimm is now known as jgrimm14:33
haypozigo, there is six.print_(), but it's uglier and only needed to support python < 2.614:33
haypo(like six.binary_type: you can directly write bytes)14:33
zigohaypo: I know already a lot about porting to py3 ! :)14:33
zigohaypo: I always feel really gulty when uploading a package to Debian without Py3 support.14:34
zigo:)14:34
haypozigo, cool. again, don't hesitate to complete the wiki page, https://wiki.openstack.org/wiki/Python3#Port_Python_2_code_to_Python_314:34
*** jungleboyj has joined #openstack-oslo14:38
zigohaypo: from dircache import listdir <--- This fails in Py3, what's the trick? :)14:43
haypozigo, what is dircache?14:43
*** SridharGaddam has quit IRC14:44
haypozigo, does it come from pysaml2?14:44
zigohaypo: It's in the standard py2.7 lib.14:44
haypoah?14:44
haypo"Deprecated since version 2.6: The dircache module has been removed in Python 3."14:44
zigohaypo: /usr/lib/python2.7/dircache.py14:44
zigohaypo: Shit, what should I do then?14:44
zigoJust re-add the code locally?14:45
zigo(it's *very* small...)14:45
haypozigo, you may use: try: from dircache import listdir except ImportError: from os import listdir14:45
zigoOh.14:45
zigoCheers! :)14:45
haypozigo, but you loose the cache14:45
haypodoes it really impact performances to list the content of a directory?14:45
haypozigo, linux probably already caches it for you14:46
zigoRight.14:46
zigoMy baby girl just woke up. Porting session has to end ! :)14:46
*** zzzeek has joined #openstack-oslo14:58
dhellmannihrachyshka: ok, sounds good. dims will be back next week, and I'm sure he'll cover those in the meeting15:00
ihrachyshkaack15:01
*** achanda has joined #openstack-oslo15:03
*** browne has joined #openstack-oslo15:05
*** achanda has quit IRC15:08
*** e0ne is now known as e0ne_15:12
*** haypo has quit IRC15:13
*** e0ne_ is now known as e0ne15:13
*** david-ly_ is now known as david-lyle15:22
*** ihrachyshka has quit IRC15:23
*** ozamiatin has quit IRC15:26
*** e0ne is now known as e0ne_15:29
*** e0ne_ is now known as e0ne15:30
*** andreykurilin__ has joined #openstack-oslo15:40
*** cdent has joined #openstack-oslo15:45
*** exploreshaifali has quit IRC15:50
*** sdake has joined #openstack-oslo15:51
*** vilobhmm1 has joined #openstack-oslo15:59
*** tsekiyama has joined #openstack-oslo15:59
*** mriedem is now known as mriedem_nom15:59
*** andreykurilin__ has quit IRC16:00
*** vilobhmm11 has joined #openstack-oslo16:01
*** ozamiatin has joined #openstack-oslo16:03
*** vilobhmm1 has quit IRC16:04
*** e0ne has quit IRC16:07
*** viktors has quit IRC16:07
*** arnaud___ has joined #openstack-oslo16:09
*** vilobhmm11 has quit IRC16:12
*** gtt116_ has quit IRC16:20
*** gtt116__ has joined #openstack-oslo16:20
*** browne has quit IRC16:20
*** ozamiatin has quit IRC16:20
*** arnaud___ has quit IRC16:39
openstackgerritJulien Danjou proposed openstack-dev/pbr: Support Unicode in requirements file  https://review.openstack.org/17735416:44
*** vilobhmm1 has joined #openstack-oslo16:47
*** vilobhmm1 has quit IRC16:47
*** vilobhmm1 has joined #openstack-oslo16:47
*** vilobhmm1 has quit IRC16:47
*** vilobhmm1 has joined #openstack-oslo16:48
*** achanda has joined #openstack-oslo16:52
*** vilobhmm1 has quit IRC16:53
*** sdake has quit IRC16:53
*** vilobhmm1 has joined #openstack-oslo16:53
*** bnemec is now known as beekneemech16:54
*** pblaho has quit IRC16:57
openstackgerritMerged openstack/oslo.concurrency: Updated from global requirements  https://review.openstack.org/17723617:01
*** browne has joined #openstack-oslo17:02
openstackgerritMoshe Levi proposed openstack/oslo.concurrency: Add timeout parameter to execute  https://review.openstack.org/17703017:07
*** exploreshaifali has joined #openstack-oslo17:08
*** mriedem_nom is now known as mriedem_rmorse17:09
*** mriedem_rmorse is now known as mriedem17:09
*** sdake has joined #openstack-oslo17:12
*** salv-orlando has quit IRC17:26
*** cdent has quit IRC17:30
*** prad_ has joined #openstack-oslo17:44
*** prad has quit IRC17:44
*** prad_ is now known as prad17:44
*** harlowja_away is now known as harlowja17:51
*** yassine has quit IRC18:06
*** jaosorior has quit IRC18:22
*** exploreshaifali has quit IRC18:25
*** salv-orlando has joined #openstack-oslo18:27
*** arnaud___ has joined #openstack-oslo18:27
*** salv-orlando has quit IRC18:30
*** ajo has quit IRC18:36
*** arnaud___ has quit IRC18:54
*** ndipanoff has quit IRC19:01
*** Ritz has joined #openstack-oslo19:15
RitzHello ! i have an architecture using oslo messaging framework where a job is submitted for processing to a server which listens on requests that come in. There are multiple nodes in the server who can work on the request. Problem is if 2 same type of job comes in for processing then i want  just one of the worker nodes to work on it. So some sort of mutex is needed.  Question is does oslo messaging support this kind of dispatche19:18
dhellmannRitz: if both workers subscribe to the same queue in rabbit, then rabbit will only deliver the message to one of them.19:21
Ritzdhellmann what i mean is if concurrently 2 requests are put in and node1 picks first and node2 picks second then they will both start working on the same type of job . I want the second message not to be picked up till first job is done by node 119:22
Ritzso some kind of lock19:23
*** tsekiyama has quit IRC19:23
dhellmannRitz: ok, our messaging library does provide that sort of feature19:28
dhellmannit sounds like what you really want is different nodes to pick up different types of jobs19:28
dhellmannor you could see about coordinating them using tooz19:28
Ritzdhellmann : actually its not different node picking up different jobs, its not allowing any node(including itself)  to pick up same type of job19:30
Ritzexample19:30
Ritzi am working on a feature called Backup , so the event type is backup19:31
Ritzso lets say i am running a db backup ...if 2 customers run the same job to backup db then there could be issues19:31
Ritzso i want to finish first backup job before starting the second job19:31
Ritzi am looking at tooz right now19:32
*** tsekiyama has joined #openstack-oslo19:33
bknudsonwhat does it take to get a stable/juno branch for pycadf?19:34
harlowjaRitz i think u want something more than a lock no? if u are having dependency chains u likely need something more19:35
harlowjau almost want http://docs.openstack.org/developer/taskflow/jobs.html (but maybe not via taskflow... not sure)19:36
Ritzi think a lock on a type of job till the listener finishes(acknowledges current job) would be sufficient19:36
harlowjawho acquires the lock?19:37
harlowjathe listener?19:37
Ritzharlowja: the first node which picked up the message should acquire the lock19:37
Ritzso if i have lets say 10 listeners then first listener should get the lock and remaining should wait in the queue19:38
harlowjaok, first node that gets the message picks it up, seems ok, but u said u want some other job to not start until after the first job finishes19:38
Ritzyes thats right19:39
*** ozamiatin has joined #openstack-oslo19:39
Ritzthe the lock is on the type of job (backup)19:39
*** prad has quit IRC19:39
*** prad has joined #openstack-oslo19:40
harlowjaso it'd be like the following19:40
harlowjameta job { backup, do-thing-after-backup, ...}19:41
harlowjaand u want to retain the ordering of that meta job right (where each part of it can be done by different entities)19:41
* harlowja just verifying i understand the problem19:41
Ritzvery sorry but i didnt understand meta job that you mentioned :(19:42
*** haypo has joined #openstack-oslo19:42
harlowjak, maybe its not needed then19:42
harlowjaso then try tooz locks, with a lock-name that is unique enough and i think u might be ok19:43
Ritzahh !19:43
*** ajo has joined #openstack-oslo19:43
Ritzi will take a look at that19:43
harlowjak19:43
harlowjaRitz http://docs.openstack.org/developer/tooz/tutorial/lock.html hopefully explains the basics (its not to hard)19:46
harlowjathe question more becomes as to what backend do u want to use (u likely need something distributed)19:46
harlowjaaka http://docs.openstack.org/developer/tooz/drivers.html19:46
harlowjachoose your poison, ha19:47
*** arnaud___ has joined #openstack-oslo19:50
*** arnaud___ has quit IRC19:54
*** jungleboyj has quit IRC19:57
Ritz:)19:57
Ritzthanks harlowja19:57
Ritzas usual !19:58
*** yassine has joined #openstack-oslo20:03
*** jungleboyj has joined #openstack-oslo20:08
*** kgiusti has left #openstack-oslo20:08
*** jgrimm is now known as zz_jgrimm20:14
haypohum, i don't understand why gate-tempest-dsvm-largeops-src-oslo.messaging fails with https://launchpad.net/bugs/1403284 : my change on requirements.txt should not impact the installed eventlet version: https://review.openstack.org/#/c/172135/7/requirements.txt20:14
openstackLaunchpad bug 1403284 in devstack "mysql (1040, 'Too many connections')" [Undecided,In progress] - Assigned to Joe Gordon (jogo)20:14
haypoi saw these tracebacks: http://pastealacon.com/3736820:14
*** yassine has quit IRC20:47
openstackgerritMoshe Levi proposed openstack/oslo.concurrency: Add timeout parameter to execute  https://review.openstack.org/17703020:56
*** jungleboyj has quit IRC20:59
*** sdake_ has joined #openstack-oslo21:04
*** andreykurilin__ has joined #openstack-oslo21:06
*** salv-orlando has joined #openstack-oslo21:07
*** sdake has quit IRC21:08
openstackgerritMerged openstack/oslo.log: Update to latest hacking  https://review.openstack.org/17049721:11
*** salv-orl_ has joined #openstack-oslo21:13
openstackgerritlifeless proposed openstack-dev/pbr: Honour new test variable PIPVERSION  https://review.openstack.org/17745721:13
*** ozamiatin has quit IRC21:14
*** salv-orlando has quit IRC21:14
*** sdake has joined #openstack-oslo21:19
*** salv-orl_ has quit IRC21:20
*** ajo has quit IRC21:21
*** haypo has quit IRC21:22
*** sdake_ has quit IRC21:23
openstackgerritMerged openstack/oslo.log: Add pypi download + version badges  https://review.openstack.org/17127621:24
*** ajo has joined #openstack-oslo21:25
*** e0ne has joined #openstack-oslo21:27
*** e0ne has quit IRC21:29
*** e0ne has joined #openstack-oslo21:33
*** ChuckC_ is now known as ChuckC21:34
*** jungleboyj has joined #openstack-oslo21:37
*** arnaud___ has joined #openstack-oslo21:39
*** arnaud___ has quit IRC21:43
*** gordc has quit IRC21:44
openstackgerritMerged openstack/oslo.messaging: Enable eventlet dependency on Python 3  https://review.openstack.org/17213521:46
*** mriedem has quit IRC21:47
*** bknudson has quit IRC21:49
*** vilobhmm11 has joined #openstack-oslo21:51
*** vilobhmm12 has joined #openstack-oslo21:53
*** stevemar has quit IRC21:53
*** vilobhmm1 has quit IRC21:53
*** vilobhmm11 has quit IRC21:56
*** prad has quit IRC21:57
*** prad has joined #openstack-oslo21:58
*** e0ne has quit IRC22:00
*** sdake_ has joined #openstack-oslo22:11
*** Kennan2 has joined #openstack-oslo22:12
*** Kennan has quit IRC22:12
*** sdake has quit IRC22:15
*** Ritz has quit IRC22:16
*** sdake has joined #openstack-oslo22:23
*** yamahata has joined #openstack-oslo22:25
*** sdake_ has quit IRC22:27
*** harlowja_ has joined #openstack-oslo22:41
*** harlowja has quit IRC22:41
*** jungleboyj has quit IRC22:44
*** ajo has quit IRC22:50
*** sdake_ has joined #openstack-oslo22:53
*** achanda has quit IRC22:53
*** stpierre has quit IRC22:56
*** sdake__ has joined #openstack-oslo22:56
*** sdake has quit IRC22:56
*** sdake_ has quit IRC22:59
*** sigmavirus24 is now known as sigmavirus24_awa23:25
*** arnaud___ has joined #openstack-oslo23:28
*** arnaud___ has quit IRC23:32
openstackgerritMerged openstack/oslo.messaging: Port ZMQ driver to Python 3  https://review.openstack.org/17439123:32
*** ChuckC has quit IRC23:36
*** takedakn has joined #openstack-oslo23:43
*** takedakn has quit IRC23:46
*** sdake__ has quit IRC23:52
*** achanda has joined #openstack-oslo23:53
*** ChuckC has joined #openstack-oslo23:56
*** achanda has quit IRC23:59
*** shardy_ has joined #openstack-oslo23:59
« Thursday, 2015-04-23 Index Saturday, 2015-04-25 »

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