Friday, 2011-09-02

vishyk looking00:14
*** jdurgin has quit IRC00:14
vishyjkoelker: almost but not quite00:18
*** misheska has quit IRC00:23
*** misheska has joined #openstack-dev00:23
*** misheska has quit IRC00:25
*** dragondm has quit IRC01:06
*** bengrue has quit IRC01:14
*** rods has quit IRC01:22
*** mdomsch has joined #openstack-dev01:29
*** dragondm has joined #openstack-dev02:05
*** dragondm has quit IRC02:11
*** jakedahn has quit IRC02:29
*** dragondm has joined #openstack-dev02:33
*** martine_ has joined #openstack-dev02:56
*** dragondm has quit IRC03:10
*** mdomsch has quit IRC03:13
*** dragondm has joined #openstack-dev03:21
*** sandywalsh_ has quit IRC04:26
*** stanchan has quit IRC05:03
*** bcwaldon has joined #openstack-dev05:09
*** novas0x2a|laptop has quit IRC05:11
*** martine_ has quit IRC05:25
*** dragondm has quit IRC05:28
*** bcwaldon has quit IRC05:50
*** banka has joined #openstack-dev06:05
ttxheckj, soren: re: "trying to come into the code of the project from no prior knowledge has been significantly challenging"06:21
ttxheckj, soren: We'll have code discovery sessions ath the design summit (sorts of tutorials given by the author of a given area) to help getting others up to speed on parts of the code.06:22
*** eday has quit IRC06:22
bankaAPI spec supports Request with MIME type versioning. Something like "application/vnd.openstack.compute-v1.1+xml".  Are such Accept headers currently accepted by nova ?06:24
*** banka has quit IRC06:53
*** reidrac has joined #openstack-dev07:02
*** banka has joined #openstack-dev07:28
*** banka has quit IRC07:29
*** chomping has quit IRC08:18
*** rods has joined #openstack-dev08:46
*** chomping has joined #openstack-dev08:53
*** deshantm_laptop has joined #openstack-dev09:04
*** darraghb has joined #openstack-dev09:08
*** alekibango has quit IRC10:57
*** sandywalsh has joined #openstack-dev11:08
*** bsza has joined #openstack-dev11:56
*** ameade has joined #openstack-dev12:18
*** lts has joined #openstack-dev12:30
*** martine_ has joined #openstack-dev12:46
zulttx: did the iso glance support make it in?12:51
ttxzul: I think so, yes... https://blueprints.launchpad.net/glance/+spec/iso-boot12:52
*** joesavak has joined #openstack-dev12:56
*** kbringard has joined #openstack-dev12:58
*** mfer has joined #openstack-dev12:59
sorenzul, ttx: Glancing at the Nova side of things, though, it seems to be a XenAPI-only thing.13:08
sorenHar har. "Glancing". Get it? It's a pun.13:08
zulsoren: oooh rimshot13:09
sandywalshsoren, congrats on the new gig13:11
sorensandywalsh: ta13:11
*** markmc has joined #openstack-dev13:15
openstackgerritDan Prince proposed a change to openstack/keystone: Update Nova and Glance paste config examples.  https://review.openstack.org/37613:19
openstackgerritDan Prince proposed a change to openstack/glance: Update paste config files with keystone examples. see ticket: lp839559  https://review.openstack.org/37713:32
*** amccabe has joined #openstack-dev13:40
*** ameade has quit IRC13:59
*** Binbin has quit IRC13:59
*** bcwaldon has joined #openstack-dev14:00
*** ameade has joined #openstack-dev14:01
*** bcwaldon_ has joined #openstack-dev14:06
*** bcwaldon_ has quit IRC14:10
*** mdomsch has joined #openstack-dev14:14
*** code_franco has joined #openstack-dev14:25
*** mfer has quit IRC14:31
*** mfer has joined #openstack-dev14:31
*** cp16net has joined #openstack-dev14:42
*** misheska has joined #openstack-dev14:50
*** deshantm_laptop_ has joined #openstack-dev14:53
*** deshantm_laptop has quit IRC14:57
*** mfer has quit IRC15:02
*** reidrac has quit IRC15:03
*** mfer has joined #openstack-dev15:20
*** nicolas2b has joined #openstack-dev15:24
*** mattray has joined #openstack-dev15:36
*** ohnoimdead has joined #openstack-dev15:40
*** deshantm_laptop_ has quit IRC15:42
openstackjenkinsProject swift build #318: SUCCESS in 31 sec: https://jenkins.openstack.org/job/swift/318/15:47
openstackjenkinsTarmac: Fixes last modified date for swift3 middleware15:47
*** nicolas2b has quit IRC15:51
*** ameade has quit IRC16:02
*** heckj has joined #openstack-dev16:25
heckjany schedule pages on the wiki or such for the design summit schedule/planning?16:27
annegentleheckj: I haven't seen any yet16:31
heckjannegentle: found http://wiki.openstack.org/Summit/Essex - wrote Stephen to see if he needed any help with setting up schedules, etc.16:31
*** jdurgin has joined #openstack-dev16:32
annegentleheckj: yep that's the place16:32
annegentleheckj: the Summit schedule will be done by ttx and the PTLs16:32
annegentleheckj: spectorclan has the program committee working on Thur/Fri schedules16:33
heckjah, OK - thanks!16:33
*** mgius has joined #openstack-dev16:52
*** relateable has quit IRC16:56
kbringardso long as we're talking about the summit17:03
kbringardsomeone was asking yesterday, but I don't recall seeing an answer17:03
kbringardif you put yourself on the design summit waiting list, should you still register for the conference?17:03
heckjkbringard - yes, they're completely separate mechanisms17:04
kbringardthe instructions tell you not to register for both, but if you don't know if you're going to get into the design summit17:04
heckjkbringard: I think the intention was originally to have them together, but it doesn't look to be working that way17:05
kbringardah, OK17:05
kbringardso then if you register for the conference but get into the design summit, your 500$ will be refunded?17:05
heckjkbringard: no idea… guessing spectorclan could say, but I'm not sure how the underpinnings are working17:06
kbringardyea, meh, I don't necessarily care since work would be paying either way, but I figured other people would want to know :-D17:06
*** markmc has quit IRC17:15
*** JordanRinke-AFK is now known as JordanRinke17:32
*** darraghb has quit IRC17:35
vishysandywalsh: ping17:41
sandywalshvishy, hey17:41
vishysandywalsh: I noticed your comment that zones are broken in diablo17:42
vishyare they irreparable?17:42
sandywalshthere were some issues from the recent refactoring and the major changes in novaclient. I think the patches have landed now that fix them.17:42
sandywalshI need to check if that includes the dabo's pending branch or not17:43
vishyok, cool17:43
sandywalshVek was doing the testing and he's at lunch now17:43
vishyif that is needed to get zones working then we need to get it in17:43
sandywalshyup17:43
vishyi don't want to ship diablo with broken zones17:43
vishy:)17:43
sandywalshabsolutely17:43
sandywalshI touch base with him when he gets back and give you an update17:43
sandywalsh*I'll17:44
comstudzones is broken in trunk17:53
comstudi'm working on coming up with a list17:53
*** mjs has joined #openstack-dev17:54
comstudat least17:54
comstudi'm getting an 'Instance None could not be found' exception17:54
comstudwhen trying to nova boot with least cost17:54
comstuderr.. s/trunk/scheduler-multifilter/ i guess17:54
*** rnorwood has joined #openstack-dev17:54
comstudcompute manager's run_instance() is being sent 'None' for instance ID17:57
*** martine_ has quit IRC17:58
*** martine has joined #openstack-dev17:58
comstudah18:00
comstudnevermind.. i apparently have the scheduler_driver commented out18:00
comstud*tries again*18:00
*** mjs has quit IRC18:01
chmouelsandywalsh: what about git tagging each version on novaclient?18:03
*** martine has quit IRC18:04
sandywalshchmouel, all for it. I think we should be more aggressive on the versioning as well. What do you think?18:06
chmouelI am all for the release often and early, I have telling a lot of people lately to pull up github/master instead of using pip18:09
chmouelsandywalsh: but what define that makes a bump in minor or  major version18:09
chmouelsandywalsh: ie: the addition of keystone should have probably be a new major version18:09
chmouelsandywalsh: or having to follow the global openstack model18:10
sandywalshmajor is marketing, minor is something that breaks compatibility (like tenant_id in url), ternary is just an update18:10
sandywalshmarketing.compatibility.patch18:11
chmouelI like that :)18:11
chmouelbut these days we have so many new features floating by18:11
chmouelwe would have to do a every day minor release18:11
sandywalshyeah, I know, that's why we haven't been so strict. Perhaps we can infer ternary from git commit # ?18:12
sandywalshhave setup be a little smarter18:12
chmouelwell I am not sure if that's very standard it may confuse people18:13
chmouel( I rem it did for me when back in the days wine project was doing that)18:13
sandywalshpyhole does this in its setup.py https://github.com/jk0/pyhole/blob/master/setup.py18:13
chmouelhumm I think the from pyhole import version in setup.py would break the pypi-install debian tools18:15
*** bengrue has joined #openstack-dev18:15
sandywalshvishy, vek tells me that trunk is good now (he applied some patches to nova and novaclient to fix issues and they've landed) ... dabo's MP needs to get freshened up before it can land.18:15
sandywalshchmouel, hmm, good point18:15
sandywalshchmouel, lemme stew on that a little ... I'd prefer to automate that stuff18:16
jk0chmouel: how would it break?18:16
chmoueljk0: well for python-cloudlb api binding I had to do some ugly stuff to make it work for it https://github.com/rackspace/python-cloudlb/blob/master/setup.py18:18
chmoueljk0: let me find the bug report18:18
chmouelsandywalsh: cool, let me know if you need some help on this18:19
vishysandywalsh: thx18:19
jk0chmouel: there shouldn't be an ImportError since the sdist is built with pyhole.version in the path18:19
jk0(unless you're referring to soemthing else018:19
comstudsandywalsh: there's a small issue with dabo's branch18:21
comstudat least one18:21
sandywalshcomstud, yup, there's conflicts, vek just commented on it18:23
comstudsandywalsh: more than that18:23
comstudyou can't build without passing a filter via API18:23
sandywalshoh, that's not good18:23
sandywalshwhy is that?18:23
comstudbuilds fail when request_spec has a filters: None18:23
comstudhttps://code.launchpad.net/~ed-leafe/nova/scheduler-multifilter/+merge/7247818:23
comstudsee last comment18:23
sandywalshah, choose_host_filter() has changed meaning18:24
*** martine has joined #openstack-dev18:25
comstudfixing and re-testing for anything else18:25
sandywalshcomstud, perhaps we should hold off on ed's branch for diablo? Trunk works now ...18:26
sandywalshcomstud, or do you think we can merge and get necessary fixes in?18:26
sandywalshcomstud, you've played with it the most I do believe18:26
chmoueljk0: just digged into it a bit more and you right it was only specific to that library: https://github.com/rackspace/python-cloudlb/issues/318:26
comstudI've not played with it until very recently again.. like today18:26
comstudand I'm using Ed's branch, not trunk18:26
comstudso I'd have to take your work ATM in saying trunk works18:27
comstuder18:27
comstudword18:27
jk0chmouel: yeah, pyhole.version doesn't depend on any 3rd party imports, so we will be fine18:27
sandywalshvek just tested it18:27
comstudok18:27
sandywalshhe filed two patches to get it going again18:27
chmoueljk0: yeah that would break on python-novaclient tho18:27
jk0what would break?18:28
comstudmaybe we should hold off on this then18:28
comstudi dunno18:28
chmoueljk0: if we were doing a from novaclient import constants; version=constants.version kind of thing18:28
sandywalshcomstud, the only reason I say so is it's not in the stories for our sprint currently and that would throw us off18:28
*** bengrue has quit IRC18:28
chmoueljk0: instead of harcoding the version in setup.py18:28
jk0chmouel: but I don't understand what would break.. when the sdist is generated, it's done from within the directory containing novaclient.constants18:29
*** bengrue has joined #openstack-dev18:30
chmoueljk0: I think  its not at the time of sdist generation but at easy_install18:30
jk0hm18:30
chmouelI amy be wrong but that's what happen on that library18:31
jk0I'll test pyhole usint easy_install18:31
chmoueli just tested and it work18:32
chmouelas it doesnt have any external dependence18:32
jk0alright, we should be fine in python-novaclient then18:32
jk0you could probably make a verbatim copy of pyhole.version in python-novaclient18:33
jk0it should "just work"18:33
chmouelah i just realise we don't have a __init__.py in novaclient/ like it used to be in cloudservers18:33
chmouelyou right it should work18:34
chmouelbecause what happen is, easy_install try to do a from novaclient import whatever, which goes to novaclient.__init__.py that does bunch of imports in there like the httplib2 which is not available as a pre-dependence since easy_install havent' seen it yet and would fail... should not happen on novaclient18:36
openstackjenkinsProject nova build #1,325: SUCCESS in 5 min 27 sec: https://jenkins.openstack.org/job/nova/1325/18:37
openstackjenkinsTarmac: Fixes a small bug which causes filters to not work at all.  Also reworks a bit of exception handling to allow the exception related to the bug to propagate up.18:37
jk0I get what you're saying18:37
jk0good point18:37
*** bcwaldon has quit IRC18:41
chmouelwhere I can find some documentaion about how nova extensions structure are made of?18:45
*** alekibango has joined #openstack-dev18:49
chmoueljk0, sandywalsh: anybody can review this https://github.com/rackspace/python-novaclient/pull/108 ?18:53
jk0I can in a few moments18:54
*** med_out is now known as medberry18:55
*** zaitcev has joined #openstack-dev18:59
*** ohnoimdead has quit IRC19:06
openstackgerritVerification of a change to openstack/keystone failed: Get Service Catalog from token - pep8  https://review.openstack.org/35919:07
openstackgerritVerification of a change to openstack/keystone failed: Adding wadl for OS-KSCATALOG extension.Fixing existing xsds.Fixing service wadls.  https://review.openstack.org/37319:09
openstackgerritA change was merged to openstack/keystone: Various documentation-related changes.  https://review.openstack.org/37519:11
openstackgerritA change was merged to openstack/keystone: Update Nova and Glance paste config examples.  https://review.openstack.org/37619:12
*** lts has quit IRC19:21
heckjRunning the unit tests from nova trunk and I'm seeing a common set of errors from them. They're all wrapped around getting non-zero exit codes from running nova-dhcpbridge: http://pastie.org/247258219:25
*** ohnoimdead has joined #openstack-dev19:25
heckjAnyone seen this, or have an idea of what's biting me? It's got to be something environment specific, but I'm at a loss as to what19:25
heckjWhen I run it manually, it returns just fine...19:26
*** markvoelker has quit IRC19:26
chmoueljk0, sandywalsh: one would object if i can make a release which has the fix-gettting-information-by-name as anne is going to release some documentation and want to reference the version that works19:34
comstudheckj: sounds dumb, but try moving your nova.conf out of the way, assuming you have one that's being pulled in19:35
heckjAhhh!19:35
* heckj tries19:35
comstudI have seen that exact error in the past on a number of occasions19:35
sandywalshchmouel, not sure of the question?19:35
annegentlesandywalsh: I need a new tarball to download19:36
comstudheckj: The 2nd idea is.. if you happen to have left any nova services running in the background... kill those off.19:36
annegentlesandywalsh: I don't wanna write instructions for the dev way19:36
sandywalshannegentle, of novaclient?19:36
sandywalshchmouel, so you want to make a bdist vs. a sdist?19:36
annegentlesandywalsh: yeah python-novaclient19:37
chmouelsandywalsh: a python setup.py register right?19:37
jk0I think they just want trunk pushed to pypi19:37
chmoueljk0: yup that's it19:37
sandywalshchmouel, I haven't tried that bdist code in a while, but sure19:37
jk0(sdist)19:37
chmouelno need a binary19:37
sandywalshah19:37
sandywalshsure, I did trunk this morning, so unless the version # changes, you'll have to delete 2.6.4 and re-add it19:38
chmouelah okay let me double check19:38
chmouelannegentle: 2.6.4 on pypi should have the fix you need19:40
chmouelsandywalsh: tks19:40
sandywalshnp19:41
heckjcomstud: Thank you! That was it - it was autoloading the conf I had from an earlier all-in-one install19:43
*** jsavak has joined #openstack-dev19:43
annegentlechmouel: thanks, getting it now!19:44
*** joesavak has quit IRC19:46
comstudheckj: Cool.  Yeah, strange error, though.  Took a while to figure that out, and I still don't know exactly why that certain non-helpful error occurs19:56
comstudheckj: But anyway, you're welcome19:56
heckjcomstud: made testing the kombu branch a right pain :-) - originally I thought it was the branch, but hadn't dug in, then finally saw it when I went back to trunk too while figuring out what I'd done differently that broke the tests19:58
*** amccabe has quit IRC19:59
*** cp16net has quit IRC20:02
*** jakedahn has joined #openstack-dev20:03
*** bcwaldon has joined #openstack-dev20:03
*** stanchan has joined #openstack-dev20:05
*** lloydbud_ has joined #openstack-dev20:06
comstudheckj: ah :)20:07
*** lloydbud_ has left #openstack-dev20:07
*** bcwaldon has quit IRC20:08
*** amccabe has joined #openstack-dev20:14
*** rnorwood has quit IRC20:25
vishyblamar: ping. is bcwaldon out for the rest of the day?20:28
*** rnorwood has joined #openstack-dev20:29
openstackgerritAdam Gandelman proposed a change to openstack/glance: Remove PWD from possible config_file_dirs.  https://review.openstack.org/37420:30
*** jsavak has quit IRC20:32
*** mdomsch has quit IRC20:51
*** code_franco has quit IRC21:06
*** code_franco has joined #openstack-dev21:07
*** vladimir3p has joined #openstack-dev21:25
vladimir3pvishy, _cerberus_: ping21:41
_cerberus_Hey21:41
vishypong21:41
vladimir3phey guys, do you know if there are any plans to add more notifications into the code21:41
_cerberus_Nothing specifically, but there's no reason not to add more21:42
vladimir3pfor example, as I see, mostly compute exceptions are wrapped and sent to notifier21:42
vladimir3pdo you know if anybody already using the notification system?21:43
vladimir3pseems like there is an option to register own drivers there ...21:43
_cerberus_We are, internally21:43
_cerberus_there is21:43
*** rnorwood has quit IRC21:44
*** nci has quit IRC21:45
vladimir3pok, thanks, will try to play with it more...21:46
vishyvladimir3p: there is also the monkey patch stuff21:47
vishyadded by nachi which puts notifications on every call21:47
vladimir3pvishy: do you mean every compute call?21:48
vishyFLAGS.monkey_patch_modules21:49
vladimir3pah, I see ... ok21:49
vishyyou define modules and it will patch every method to send a notification21:49
vladimir3pvishy: this is very interesting ... I missed it before. It can probably potentially overflow the notification system/hub... and it will depend on driver or hub to filter them efficiently21:53
*** heckj has quit IRC21:56
*** Gordonz has quit IRC21:56
*** rnorwood has joined #openstack-dev21:57
*** code_franco has quit IRC22:00
vishyyes22:07
vishyvladimir3p: that is why it is turned off by default :)22:07
*** kbringard has quit IRC22:13
*** kbringard has joined #openstack-dev22:14
*** dragondm has joined #openstack-dev22:23
openstackjenkinsProject nova build #1,326: SUCCESS in 3 min 35 sec: https://jenkins.openstack.org/job/nova/1326/22:25
openstackjenkinsTarmac: Added list of security groups to the newly added extension (Createserverext)  for the Create Server and Get Server detail responses.22:25
*** rnorwood has quit IRC22:29
*** vladimir3p has quit IRC22:35
*** vladimir3p has joined #openstack-dev22:37
*** amccabe has quit IRC22:39
*** mattray has quit IRC22:43
*** mfer has left #openstack-dev22:45
*** bengrue has quit IRC22:46
*** bsza has quit IRC22:58
*** bsza has joined #openstack-dev22:59
*** bsza has quit IRC23:10
*** dragondm has quit IRC23:22
*** lorin1 has joined #openstack-dev23:31
*** lorin1 has left #openstack-dev23:33
*** nci has joined #openstack-dev23:39
*** vladimir3p has quit IRC23:56
*** dragondm has joined #openstack-dev23:56

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