Wednesday, 2014-05-21

*** oz_akan_ has joined #openstack-marconi00:09
*** jergerber has quit IRC00:11
*** rossk has quit IRC00:58
*** oz_akan_ has quit IRC01:01
*** oz_akan_ has joined #openstack-marconi01:01
*** malini_afk is now known as malini01:17
*** flwang has quit IRC01:44
*** shakamunyi has quit IRC01:45
*** shakamunyi has joined #openstack-marconi01:59
*** oz_akan_ has quit IRC02:00
*** oz_akan_ has joined #openstack-marconi02:00
*** oz_akan_ has quit IRC02:00
*** shakamunyi has quit IRC02:05
*** malini is now known as malini_afk02:05
*** malini_afk is now known as malini02:09
*** prashanthr_ has joined #openstack-marconi02:17
*** prashanthr_ has quit IRC02:39
*** rwsu has quit IRC02:44
*** balajiiyer has joined #openstack-marconi02:47
*** prashanthr_ has joined #openstack-marconi03:00
*** oz_akan_ has joined #openstack-marconi03:01
*** oz_akan_ has quit IRC03:06
*** oz_akan_ has joined #openstack-marconi03:09
*** oz_akan_ has quit IRC03:13
*** balajiiyer has quit IRC03:14
*** cpallares has quit IRC03:20
*** oz_akan_ has joined #openstack-marconi03:23
*** vkmc has quit IRC03:25
*** oz_akan_ has quit IRC03:44
*** nosnos has joined #openstack-marconi03:48
*** malini is now known as malini_afk03:55
*** prashanthr_ has quit IRC03:59
*** prashanthr_ has joined #openstack-marconi04:17
*** haomaiwang has joined #openstack-marconi04:23
*** prashanthr_ has quit IRC04:34
*** oz_akan_ has joined #openstack-marconi04:45
*** oz_akan_ has quit IRC04:49
*** igor_ has quit IRC05:02
*** igor has joined #openstack-marconi05:06
*** reed has quit IRC05:21
*** prashanthr_ has joined #openstack-marconi05:39
*** oz_akan_ has joined #openstack-marconi05:45
*** prashanthr_ has quit IRC05:47
*** oz_akan_ has quit IRC05:50
*** AAzza has joined #openstack-marconi06:02
*** prashanthr_ has joined #openstack-marconi06:04
*** oz_akan_ has joined #openstack-marconi06:09
*** oz_akan_ has quit IRC06:13
*** flwang has joined #openstack-marconi06:16
*** prashanthr_ has quit IRC06:43
*** prashanthr_ has joined #openstack-marconi06:59
*** flaper87|afk is now known as flaper8707:02
*** oz_akan_ has joined #openstack-marconi07:10
*** oz_akan_ has quit IRC07:14
*** prashanthr_ has quit IRC07:18
*** prashanthr_ has joined #openstack-marconi07:32
*** prashanthr_ has quit IRC08:01
*** oz_akan_ has joined #openstack-marconi08:10
*** oz_akan__ has joined #openstack-marconi08:12
*** oz_akan_ has quit IRC08:15
*** oz_akan__ has quit IRC08:17
*** haomaiwang has quit IRC08:30
*** haomaiwang has joined #openstack-marconi08:30
*** jamie_h has joined #openstack-marconi08:31
*** haomaiw__ has joined #openstack-marconi08:33
*** haomaiwang has quit IRC08:35
*** prashanthr_ has joined #openstack-marconi08:43
*** prashanthr_ has quit IRC09:02
*** oz_akan_ has joined #openstack-marconi09:13
*** oz_akan_ has quit IRC09:17
*** flwang has quit IRC09:25
*** ykaplan has joined #openstack-marconi09:56
*** oz_akan_ has joined #openstack-marconi10:14
*** haomaiw__ has quit IRC10:15
*** haomaiwang has joined #openstack-marconi10:16
*** oz_akan_ has quit IRC10:18
*** haomaiw__ has joined #openstack-marconi10:23
*** haomaiwang has quit IRC10:26
*** oz_akan_ has joined #openstack-marconi11:14
*** nosnos has quit IRC11:16
*** oz_akan_ has quit IRC11:19
*** mwagner_lap has quit IRC11:33
*** tedross has joined #openstack-marconi11:38
*** vkmc has joined #openstack-marconi11:52
*** vkmc has quit IRC11:52
*** vkmc has joined #openstack-marconi11:52
*** oz_akan_ has joined #openstack-marconi12:15
*** oz_akan_ has quit IRC12:17
*** oz_akan_ has joined #openstack-marconi12:18
*** oz_akan_ has quit IRC12:26
*** flaper87 is now known as flaper87|afk12:38
*** sriram has joined #openstack-marconi12:46
*** sriram has quit IRC12:47
*** sriram has joined #openstack-marconi12:48
*** mwagner_lap has joined #openstack-marconi12:50
*** abettadapur has joined #openstack-marconi12:54
*** prashanthr_ has joined #openstack-marconi12:55
*** jchai has joined #openstack-marconi12:59
*** mpanetta has joined #openstack-marconi13:22
*** oz_akan_ has joined #openstack-marconi13:30
*** balajiiyer has joined #openstack-marconi13:40
*** cpallares has joined #openstack-marconi14:01
*** Obulpathi has joined #openstack-marconi14:03
*** Obulpathi has quit IRC14:04
*** Obulpathi has joined #openstack-marconi14:05
*** rwsu has joined #openstack-marconi14:15
*** cath6184 has joined #openstack-marconi14:22
*** cath6184 has quit IRC14:26
*** shakamunyi has joined #openstack-marconi14:31
*** kgriffs|afk is now known as kgriffs14:32
*** haomaiw__ has quit IRC14:37
*** haomaiwa_ has joined #openstack-marconi14:37
*** haomaiw__ has joined #openstack-marconi14:39
*** haomaiwa_ has quit IRC14:42
abettadapurkgriffs: ping14:43
kgriffsabettadapur: pong14:44
abettadapurkgriffs: so i am working on making sure headers are present for requests14:44
abettadapurkgriffs: i implemented it as a before hook, but that made all the v1 stuff also require the headers14:44
abettadapurkgriffs: would it just be better to check in the on_x methods themselves?14:45
kgriffslet me see...14:46
kgriffsyes, those changes should be made under transport/wsgi/v1_114:46
abettadapurkgriffs: alright, thanks14:47
abettadapurkgriffs: basically just if project_id is none: raise exception ?14:47
kgriffsright, you can raise a bad request exception14:47
abettadapurawesome14:48
abettadapurthanks14:48
kgriffsyou'll also need to make sure each change has a test14:48
kgriffslet me see where those would go14:48
abettadapuri was just adding the proper headers to the existing v1.1 tests...14:48
kgriffsthe functional tests?14:49
abettadapurunit tests, haven't gotten to func14:49
kgriffsoic14:49
kgriffsI think it would be good to also do negative tests14:49
kgriffsensure that the correct error status is returned if project id is missing, for example14:50
abettadapurok14:50
abettadapurlots of changes :D14:50
kgriffsthanks for tackling this, btw!14:51
abettadapurno problem14:51
*** jmckind has joined #openstack-marconi14:56
*** LaceyKite has joined #openstack-marconi15:03
openstackgerritNataliia Uvarova proposed a change to openstack/marconi: Add separate requirements file for Python 3  https://review.openstack.org/9464215:05
*** flwang has joined #openstack-marconi15:07
Alex_Gaynorflaper87|afk: related to ^, not sure if you saw, but a bunch of py3k patches landed into python-memcached's github15:08
kgriffsAlex_Gaynor: excellent!15:09
*** jmckind has quit IRC15:09
*** jchai is now known as jchai_afk15:10
kgriffsAlex_Gaynor: different topic, but I noticed the pypy gate on Marconi started failing a few weeks ago. I can't tell whether it is a bug in pypy or problem with a library we depend on...15:10
kgriffs"Fatal RPython error: KeyError"15:10
kgriffshttp://logs.openstack.org/02/90202/3/check/gate-marconi-pypy/727542c/15:10
Alex_Gaynorkgriffs: hmm, this looks like a bug that was fixed in pypy 2.3 (which came out last week), I wonder if there's a way to see what pypy version this had15:11
kgriffsah, I can try out 2.3 on my local box. if that works, we need them to update in the gate15:12
prashanthr_kgriffs: good morning :)15:14
Alex_Gaynorkgriffs: it'll get updated as soon as new images are published; it's in the PPA15:15
kgriffsrock on15:15
kgriffsprashanthr_: howdy!15:15
prashanthr_kgriffs : :) . I had a couple of doubts. can i ask them now ?15:16
kgriffsgo for it15:16
prashanthr_kgriffs :1. Redis supports SSL connections from the client to the redis server. Do we need to support this configuration as well ?15:17
kgriffsyes15:17
kgriffsIf it were up to me, we would TLS all the things, but hypervisors don't always expose AES_NI (and lots of legacy servers out there still have old CPUs that don't support it anyway)15:18
kgriffseven so, I think is many cases performance is negligible once the connection is established, so I don't see any reason not to do it for long-lived connections to the data store15:19
kgriffsanyway, that's my $0.0215:19
*** malini_afk is now known as malini15:19
*** jchai_afk is now known as jchai15:19
prashanthr_kgriffs:  I shall add support to SSL as well. thanks :)15:19
prashanthr_2. Since i am developing on a seperate github project. How can i get my code reviewed ?15:19
kgriffsprashanthr_: btw, make sure to enable cert verification, some libs don't do that by default15:21
prashanthr_kgriffs: Sure. will check if the redis client supports it.15:21
kgriffs2. mmm, good question.15:22
kgriffsyou could fork and then do pull requests and have us review15:22
kgriffsor... you could just request in here for us to look at new versions of the code...15:22
kgriffsin the case of redis, I suspect we will want to pull this into the main openstack/marconi repo so there will be a review on the entire driver when that happens15:23
kgriffsjust some ideas off the top of my head15:23
kgriffsif you go the PR route, you can manually rebase and merge via fast-forward to avoid a bunch of extraneous "merge" commits15:25
*** reed has joined #openstack-marconi15:25
kgriffsyou may also want to set up Travis CI and coveralls to gate your commits15:25
prashanthr_kgriffs: I think i will start working with a fork for now.  Will get the base driver up and then try to merge the into the main marconi repo.15:28
kgriffskk15:29
*** igor has quit IRC15:50
*** igor__ has joined #openstack-marconi15:51
*** prashanthr_ has quit IRC15:58
kgriffsAlex_Gaynor: BTW, I tested PyPy 2.3 locally and it is passing now. w00t16:05
Alex_Gaynorkgriffs: great!16:05
*** alcabrera|afk is now known as alcabrera16:07
malinikgriffs, flaper87|afk: since we are not obsoleting metadata, does it make sense to rbring back https://review.openstack.org/#/c/76093/ ?16:09
kgriffsperhaps16:16
kgriffsif people don't want to lose their metadata but still truncate the queue, then bringing this patch back but requiring a query param may be useful...16:17
kgriffshowever, I'm still on the fence. this patch isn't critical and makes the API more complex - may not be worth it16:17
kgriffsa user can always read their metadata, delete the queue, then create a new one with the same metadata16:18
*** alcabrera is now known as alcabrera|afk16:20
*** alcabrera|afk is now known as alcabrera16:24
*** whenry has joined #openstack-marconi16:29
malinikgriffs: makes senseā€¦it is easy to accidentally delete all messages if this patch lands16:30
kgriffsI went ahead and commented on the patch16:33
malinisaw tht..thanks!16:35
*** balajiiyer has quit IRC16:36
abettadapurmalini: WRT the json schema16:43
maliniabettadapur: thanks for tht patch!!!16:43
abettadapurthe hrefs you highlighted are different from each other16:43
abettadapurshould i pull them out as vars still?16:44
abettadapur(i pulled out the message object as one big var)16:44
maliniyou mean message_href vs claim_href ?16:44
abettadapurso one of them matches /v1/queues/.../messages?ids=...16:44
abettadapurthe other matches /v1/queues/.../messages/...16:45
abettadapurthe third matches v1/queues/.../messages/...?claim_id=...16:45
abettadapuris it still worth it to extract them?16:45
malinino, unless we are using the same href definitions in multiple places?16:46
abettadapurthe only place was within the big message object16:46
abettadapurwhich i extracted16:46
*** rossk has joined #openstack-marconi16:47
abettadapurunless we are talking about across files...16:47
kgriffsbe back later (lunch)16:48
*** kgriffs is now known as kgriffs|afk16:48
malinicross files? we use the schema only in response.py -rt?16:51
abettadapuryeah, ok16:55
abettadapurnever mind16:55
abettadapurone other thing. so the skipped test appears to be a bug16:55
abettadapurwhere do we report these?16:55
maliniabettadapur: https://bugs.launchpad.net/marconi16:56
maliniPlz refer to the bug ID, in the message for skipped tests16:56
abettadapurok16:56
malini we should just skip the schema validation for these tests - we might be missing other bugs, by skipping the tests altogtther16:57
*** balajiiyer1 has joined #openstack-marconi17:00
abettadapurok, just comment out the line then?17:00
malini& add a comment above tht line with bug #ID & details17:00
abettadapurawesome17:02
*** mkoderer has quit IRC17:02
*** haomaiw__ has quit IRC17:02
*** mpanetta_ has joined #openstack-marconi17:04
*** mpanetta_ has quit IRC17:05
*** mpanetta_ has joined #openstack-marconi17:05
*** mpanetta has quit IRC17:06
*** mpanetta_ is now known as mpanetta17:07
*** abettadapur has quit IRC17:07
*** LaceyKite has quit IRC17:09
*** oz_akan_ has quit IRC17:16
*** oz_akan_ has joined #openstack-marconi17:17
*** abettadapur has joined #openstack-marconi17:19
openstackgerritAlex Bettadapur proposed a change to openstack/marconi: V1 Tests JsonSchema  https://review.openstack.org/9421217:38
peoplemergeflaper87|afk: I need to triage a couple of env issues here before starting to tackle fun riak stuff, please excuse some radio silence :)17:51
notmynamekgriffs|afk: thanks for jumping in on the keystone token size thread18:07
*** LaceyKite has joined #openstack-marconi18:08
*** abettadapur has quit IRC18:20
*** igor__ has quit IRC18:34
*** igor_ has joined #openstack-marconi18:35
*** abettadapur has joined #openstack-marconi18:46
vkmchey guys, I have some doubts regarding this bug https://bugs.launchpad.net/marconi/+bug/1243752... which error code should we return?18:51
vkmcif we consider that we create queues in a lazy way18:52
maliniI think sriram did consider some of tht18:52
vkmcand sending messages to a queue that doesn't exist is ok18:52
vkmcthen retrieving messages for a queue that doesn't exist should be ok too18:52
vkmcmalini, oh cool :)18:53
maliniI wonder if the bug is still relevant when sriram's  changes get merged?18:53
malinisriram: ^18:53
*** shakamunyi has quit IRC18:57
*** kgriffs|afk is now known as kgriffs19:02
kgriffsnotmyname: sure thing.19:03
* alcabrera catches up19:03
alcabreravkmc: if the decision stands to pursue lazy queue creation (queues as topics), then 204 is the right way to for sure19:08
alcabreralikewise for retrieving from an empty topic. :)19:09
vkmcalcabrera, great :) so, once sriram's changes get merged, it's no longer a bug19:10
alcabrerayup. :)19:11
*** balajiiyer1 has left #openstack-marconi19:11
vkmcalcabrera, and what about v1.0?19:15
alcabrerahmmm19:15
alcabrerain v1.019:16
alcabrerawe're assuming the presence of strict queues19:16
alcabreraso a 404 would be the way to go19:16
alcabreravkmc: ^19:16
sriramyeah you guys are right.. sorry, I was away before.19:17
srirammalini: we need to open a blueprint for the unit test stuff I was talking about.19:18
sriramunit test decoupling between v1 and v1.119:18
malinisriram: can you create one, plz?19:18
vkmcalcabrera, cool... so, should this be considered a backport?19:18
srirammalini: allright. let me do that.19:19
alcabreravkmc: hmm, sort of? I'm torn on "backport", heh. >.>19:21
alcabreraI feel ike v1.0 is the current stable release19:21
alcabreraand v1.1 is still beta - though I could be wrong19:21
alcabrerabeta as in - not yet feature complete19:21
alcabreraso given that v1.0 is current, it'd just be a fix, w/o the backport19:22
srirammalini: https://blueprints.launchpad.net/marconi/+spec/decoupling-unit-tests19:23
maliniI wish we can get this done by j119:24
malinibut we already have a lot of stuff19:25
sriramthis would need to be done, if we intend on changing any functionality. adding new tests still work.19:25
maliniI made it J119:26
malinitht is by June-12 th19:26
*** whenry has quit IRC19:27
*** whenry has joined #openstack-marconi19:27
vkmcalcabrera, I see... :) well, I'll keep digging into it then19:33
alcabrerathanks, vkmc!19:34
*** LaceyKite has quit IRC19:35
malinikgriffs: why did we change the href format in 1.1 https://wiki.openstack.org/wiki/Marconi/specs/api/v1.1#Get_a_Specific_Message ?19:39
kgriffslooking19:40
kgriffsmalini: i think that was done to simply save some bytes in the response message. is it causing a problem?19:45
malinino..was just curious19:46
maliniwhy do we return href , if its the same as id?19:46
maliniI mean is there any value keeping it around?19:46
malini& it gets different https://wiki.openstack.org/wiki/Marconi/specs/api/v1.1#Get_a_Set_of_Messages_by_ID19:47
*** ametts has joined #openstack-marconi19:51
kgriffsfirst off, the href is supposed to be opaque. It just happens to be the same as the ID in the example, but it may not be.19:54
kgriffssecond, when getting a list of messages you have to include the "messages/" prefix due to the rules for resolving a relative URI19:54
kgriffsgiven19:54
kgriffs /foo/bar19:55
kgriffsand a relative URI of "123"19:55
kgriffscombining them gives19:55
kgriffs /foo/12319:55
kgriffs(I know, not what you might expect)19:55
kgriffsso you have to give19:55
kgriffsbar/12319:56
kgriffsto get19:56
kgriffs /foo/bar/12319:56
kgriffsmake sense?19:56
kgriffsmalini: however, I would like to propose at our next team meeting removing the "Get a Specific Message" endpoint from the v1.1 api19:57
malinitht sounds sane19:57
kgriffsit doesn't serve any useful purpose that I can see, and it has led to confusion in the community19:57
*** jchai is now known as jchai_afk20:00
vkmcIs anybody else getting this output when running tests? http://paste.openstack.org/show/81090/20:13
*** sriram has quit IRC20:22
kgriffsvkmc: I haven't seen that myself20:25
kgriffsthough, I just run my tests through tox usually20:25
vkmckgriffs, hmm maybe is my local env,... I tried running tox too and I'm getting a different error http://paste.openstack.org/show/81095/20:27
vkmckgriffs, I remember this was an incompatibility problem with the tox version, maybe I have an outdated version20:28
AAzzavkmc: use tox==1.6.120:28
AAzzavkmc: you have too new version i think, for the last error20:28
vkmcAAzza, thanks :) I do yeah, v1.7.120:29
kgriffsah, right. I'd forgotten about that20:30
kgriffsyeah, you have to downgrade20:30
alcabreragtk20:32
alcabreraI had no idea about that20:33
alcabrerathanks, AAzza!20:33
* vkmc running tests now \o/20:34
*** abettadapur has quit IRC20:39
alcabrerayay!20:48
alcabreraI'm out for the night.20:56
alcabreratake care, all!20:56
*** fifieldt has joined #openstack-marconi20:57
mpanettaGnight!20:57
*** jchai_afk is now known as jchai21:00
vkmcgood night! take care o/21:01
*** oz_akan_ has quit IRC21:01
*** AAzza has quit IRC21:02
*** alcabrera is now known as alcabrera|afk21:05
openstackgerritA change was merged to openstack/marconi: Updated from global requirements  https://review.openstack.org/9315521:06
openstackgerritA change was merged to openstack/marconi: Revert "Disable Metadata write operations on v1.1"  https://review.openstack.org/9437421:08
*** tedross has quit IRC21:11
*** balajiiyer has joined #openstack-marconi21:12
*** balajiiyer has quit IRC21:13
*** jchai is now known as jchai_afk21:14
*** jchai_afk is now known as jchai21:18
*** reed has quit IRC21:18
*** reed has joined #openstack-marconi21:18
*** balajiiyer1 has joined #openstack-marconi21:34
*** balajiiyer1 has left #openstack-marconi21:39
*** jamie_h has quit IRC21:42
*** jchai has quit IRC21:47
*** mwagner_lap has quit IRC21:47
*** ykaplan has quit IRC21:55
tjanczuk_hello, anyone there?21:59
*** kgriffs is now known as kgriffs|afk22:03
*** mpanetta has quit IRC22:06
*** malini is now known as malini_afk22:07
*** oz_akan_ has joined #openstack-marconi22:16
*** Obulpathi has quit IRC22:22
*** oz_akan_ has quit IRC22:22
*** notmyname_ has joined #openstack-marconi22:33
*** notmyname has quit IRC22:33
*** notmyname_ is now known as notmyname22:34
*** kgriffs|afk has quit IRC22:34
*** igor__ has joined #openstack-marconi22:34
*** kgriffs|afk has joined #openstack-marconi22:36
*** kgriffs|afk is now known as kgriffs22:36
*** ChanServ sets mode: +o kgriffs22:36
*** reed has quit IRC22:38
*** igor_ has quit IRC22:38
*** whenry has quit IRC22:39
*** reed has joined #openstack-marconi22:39
*** rwsu has quit IRC22:49
*** whenry has joined #openstack-marconi22:55
*** vkmc has quit IRC22:59
*** nosnos has joined #openstack-marconi23:10
*** whenry has quit IRC23:11
*** rwsu has joined #openstack-marconi23:12
*** mwagner_lap has joined #openstack-marconi23:14
*** anamalagon has joined #openstack-marconi23:25
*** anamalagon is now known as amalagon23:26
openstackgerritAlex Gaynor proposed a change to openstack/marconi: Fixed a few typos in the codebase  https://review.openstack.org/9474923:40
*** openstackgerrit has quit IRC23:49
*** openstackgerrit has joined #openstack-marconi23:51

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