Tuesday, 2014-03-04

*** lakspace has quit IRC00:10
*** lakspace has joined #openstack-marconi00:12
*** anamalagon has joined #openstack-marconi00:19
*** rwsu has quit IRC00:27
*** rwsu has joined #openstack-marconi00:30
*** anamalagon has quit IRC00:38
*** openstackgerrit has quit IRC00:44
*** openstackgerrit has joined #openstack-marconi00:44
*** flwang has quit IRC00:56
*** cpallares has quit IRC01:02
*** rossk has quit IRC01:08
*** rwsu has quit IRC01:17
*** nosnos has joined #openstack-marconi01:29
*** ayoung_afk is now known as ayoung01:30
*** malini_afk is now known as malini02:06
*** flwang has joined #openstack-marconi02:47
*** malini is now known as malini_afk02:51
*** nidhi has joined #openstack-marconi03:14
*** nidhi has joined #openstack-marconi03:24
*** nidhi has quit IRC03:32
*** chandan_kumar has joined #openstack-marconi03:37
openstackgerritAngus Salkeld proposed a change to openstack/python-marconiclient: Add support for claims  https://review.openstack.org/7261404:08
*** chandan_kumar has quit IRC04:53
*** nidhi has joined #openstack-marconi05:05
*** nidhi has quit IRC05:35
*** flwang has quit IRC06:58
*** saju_m has joined #openstack-marconi07:12
*** ayoung is now known as ayoung-ZZzZzzZz_07:12
*** flwang has joined #openstack-marconi07:17
*** nosnos has quit IRC07:31
*** nosnos_ has joined #openstack-marconi07:31
openstackgerritFengqian Gao proposed a change to openstack/marconi: Sort requirement files in alphabetical order  https://review.openstack.org/7707807:44
*** flaper87|afk is now known as flaper8707:58
flwangflaper87: morning08:02
*** nosnos_ has quit IRC08:04
flaper87flwang: yo yo, how are you doing?08:04
flwangflaper87: I guess you're drinking coffee now, haha08:04
*** nosnos has joined #openstack-marconi08:04
flwangflaper87: some questions are coming...08:05
flaper87flwang: TONS of coffee08:06
flaper87flwang: shoot08:06
flwangflaper87: i'm working on the /health bp, adding some test cases08:08
flwangflaper87: 1. you know, sqlite is deprecating, and the default test db backend is sqlite, and my test case is trying to get active shards number, so any idea about this?08:09
flwang2. any easy way to enable the mongodb as unit test backend instead of sqlite?08:10
flwang3. seems we have not completed the functional test against v1.1, did i miss anything?08:11
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Deprecate sqlite in favor of sqlalchemy  https://review.openstack.org/7648408:16
*** saju_m has quit IRC08:26
flaper87flwang: that patch ^ was approed yday but failed to merge08:27
flaper87I'll ninja approve it now08:27
flaper87with that you'll be able to switch to sqlalchemy08:27
flaper87flwang: re mongodb, you can use a config file that enables mongodb08:27
flwang#1 cool08:30
flwang#2 wsgi_mongodb.conf?08:32
*** flaper87|afk has joined #openstack-marconi08:35
*** flaper87|afk has quit IRC08:50
*** flwang has quit IRC08:54
*** haomaiwang has quit IRC08:57
*** haomaiwang has joined #openstack-marconi09:00
*** flwang has joined #openstack-marconi09:00
*** whenry has quit IRC09:05
*** sballe has quit IRC09:08
*** sballe has joined #openstack-marconi09:08
openstackgerritA change was merged to openstack/marconi: Deprecate sqlite in favor of sqlalchemy  https://review.openstack.org/7648409:10
*** malini_afk is now known as malini09:11
flaper87flwang: ^^^^^09:12
*** sballe has quit IRC09:17
*** flwang has quit IRC09:24
seiflotfyflaper87 how do i run marconi tests09:30
flaper87seiflotfy: python setup.py testr09:32
flaper87seiflotfy: the r at the end is not a typo09:32
*** ykaplan has joined #openstack-marconi09:36
*** saju_m has joined #openstack-marconi09:41
seiflotfyflaper87 python setup.py testr09:47
seiflotfyusage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]09:47
seiflotfy   or: setup.py --help [cmd1 cmd2 ...]09:47
seiflotfy   or: setup.py --help-commands09:47
seiflotfy   or: setup.py cmd --help09:47
seiflotfyerror: invalid command 'testr'09:47
flaper87seiflotfy: did you installed test-requirements ?09:49
flaper87seiflotfy: $ pip install -r test-requirements.txt09:49
seiflotfyyeah just noticed i forgot to do that09:49
seiflotfyim on the wrong computer (on a mac)09:49
seiflotfyok works09:50
seiflotfyfalse alarm09:50
seiflotfyok fixing time09:50
flaper87mmh, it should work on mac too10:08
flaper87that's how cool marconi is10:08
seiflotfyflaper87 MismatchError: 204 != 20110:27
seiflotfyi keep getting this10:27
*** flwang has joined #openstack-marconi10:47
*** jamieh has joined #openstack-marconi10:49
flwangseiflotfy: ping10:49
*** haomaiwang has quit IRC11:08
*** haomaiwang has joined #openstack-marconi11:09
*** haomaiwang has quit IRC11:14
seiflotfyflwang hacking11:15
*** ykaplan has quit IRC11:25
seiflotfyflwang flaper87 oen of the tests return a 204 instead of a 20111:32
seiflotfyflaper87 how bad is this then11:33
flaper87seiflotfy: what test is failing?11:46
seiflotfyFAIL: tests.functional.wsgi.v1.test_queues.TestInsertQueue.test_insert_queue_header_asterisk11:47
seiflotfytags: worker-111:47
seiflotfyEmpty attachments:11:47
seiflotfy  pythonlogging:'marconi'11:47
seiflotfy  stderr11:47
*** balajiiyer has joined #openstack-marconi11:50
seiflotfyflaper87 generally the test suite seems to be broken11:54
seiflotfyfor me11:54
flaper87is that the only test failing ? Try running it with concurrency 111:55
flaper87seiflotfy: python setup.py testr --slowest --testr-args='--concurrency 1'11:55
flaper87seiflotfy: and pls, use http://paste.openstack.org/ for logs :)11:55
seiflotfysorry was on it11:56
seiflotfysenior flaper8711:57
seiflotfyhere is a new one11:57
flaper87seiflotfy: you're using latest and you cleaned all the .pyc files right ?12:01
flaper87the test pass for me12:01
seiflotfyflaper87 r u testing using mongodb12:04
flaper87seiflotfy: nope12:05
seiflotfywhat r u using12:05
flaper87seiflotfy: yes, which is the default12:05
seiflotfyyeah seems to break with mongodb12:05
flaper87does that error come up when testing with mongo?12:06
flaper87lemme try12:06
seiflotfyhow does the conf look like for sqlite12:06
flaper87FWIW we're working on gating with mongo12:06
flaper87no conf needed, it's the default12:06
*** ekarlso has quit IRC12:09
seiflotfyyeah using the default conf results in big issues here when running the tests12:13
seiflotfyseems like it cant connect to 'sqlite://memory'12:15
flaper87seiflotfy: it's 'sqlite://:memory:'12:16
seiflotfyah yes12:16
seiflotfydo i need to set that up in the conf12:16
flaper87you shouldn't need that, if you've the latest code, that should be the default value12:17
seiflotfyfrom git12:18
*** vkmc has joined #openstack-marconi12:18
*** vkmc has quit IRC12:18
*** vkmc has joined #openstack-marconi12:18
seiflotfyit says12:18
seiflotfy Storage driver module (e.g., mongodb, sqlite)12:18
seiflotfystorage = mongodb12:18
*** balajiiyer has quit IRC12:21
seiflotfyflaper87 http://paste.openstack.org/show/72139/12:23
seiflotfyhere u go12:23
*** malini is now known as malini_afk12:28
flaper87seiflotfy: ooooooooooooopssssssssssssssssssssssssssssssssssssssssssssss12:39
flaper87seiflotfy: looks like that didn't get replaced in the last patch12:39
seiflotfycan i do that?12:39
seiflotfyi can fix it i guess12:39
flaper87seiflotfy: sure you can12:39
flaper87and thanks a lot12:39
flaper87seiflotfy: just replace it here: tests/unit/common/storage/test_utils.py12:40
flaper87I mean, s/memory/:memory:/12:40
*** saju_m has quit IRC12:40
seiflotfyon it12:40
seiflotfyshould i file a bug or just create a gerrit pull request12:40
*** mwagner_lap has quit IRC12:40
flaper87seiflotfy: you should but not need this time12:41
seiflotfyyeah i am coming from cinder to dedicate my time here12:42
seiflotfylove cinder but this is badass :D12:42
flaper87seiflotfy: AWESOME! :D12:42
seiflotfylove the specs and everything you guys did12:42
flaper87I'm glad you like it12:42
*** ykaplan has joined #openstack-marconi12:42
*** saju_m has joined #openstack-marconi12:51
seiflotfyhow do i test a specific test case12:53
flaper87seiflotfy: you can either use nosetests (you'll need to install nose) or just pass the test name to the testr command12:54
flaper87seiflotfy: python setup.py testr --slowest --testr-args='--concurrency 1 test.unit.common.....'12:54
flaper87if you want do use pdb, then nosetests -s is what you need12:54
flaper87(or whatever testing tool you prefer)12:54
seiflotfyyeah seems like its stil lfailing12:57
seiflotfythis is interesting12:57
flaper87seiflotfy: git diff ?12:57
flaper87seiflotfy: and that test still fails ?13:00
flaper87that looks correct13:01
seiflotfy  File "tests/unit/common/storage/test_utils.py", line 30, in test_can_connect_suceeds_if_good_uri_sqlite13:01
seiflotfy    self.assertTrue(utils.can_connect('sqlite://:memory:'))13:01
seiflotfy  File "/usr/local/Cellar/python/2.7.6/Frameworks/Python.framework/Versions/2.7/lib/python2.7/unittest/case.py", line 424, in assertTrue13:01
seiflotfy    raise self.failureException(msg)13:01
seiflotfyi need my linux again13:01
seiflotfyi used to pastebin with one command13:01
flaper87seiflotfy: linux forever :D13:01
*** saju_m has quit IRC13:04
seiflotfyhttp://paste.openstack.org/show/72151/ flaper8713:07
*** flaper87 has quit IRC13:17
*** flaper87 has joined #openstack-marconi13:17
*** barjavel.freenode.net sets mode: +o flaper8713:17
*** ykaplan has quit IRC13:18
*** ykaplan has joined #openstack-marconi13:18
*** ekarlso has joined #openstack-marconi13:23
openstackgerritFei Long Wang proposed a change to openstack/marconi: feat(v1.1): Implement new /health endpoint for v1.1  https://review.openstack.org/7655513:24
flwangflaper87: ping13:26
flaper87flwang: reviewing13:26
flwangflaper87: haha, great, thanks, just want to know if the overall shape is what you want to see :)13:26
flaper87flwang: yes, it's what I had in mind :D13:27
flwangflaper87: i'm trying to add more test cases13:27
flwangflaper87: any suggestion?13:27
flaper87flwang: I've some comments, comming13:30
flwangflaper87: cool13:30
flwangflaper87: thanks, guru13:30
*** sballe has joined #openstack-marconi13:33
*** tedross has joined #openstack-marconi13:44
flwangflaper87: ping13:48
flwangflaper87: seems there is no check if the [drivers] storage = sqlite  NOT compatiable with [drivers:storage:sqlalchemy], did i miss something?13:49
*** nosnos has quit IRC13:56
*** ykaplan has quit IRC14:00
*** balajiiyer has joined #openstack-marconi14:02
*** alcabrera has joined #openstack-marconi14:04
alcabreraGood morning, all. :)14:05
seiflotfyflwang same thing here14:06
flwangalcabrera: hey, morning14:06
seiflotfyflwang i am trying to deal with that issue too14:07
flwangseiflotfy: what issue?14:07
seiflotfyusing the default conf with sqlite as a backend does not work14:07
flwangseiflotfy: really? not my case14:09
seiflotfyok i cant seem to get it running at all14:09
seiflotfycan you please shoe me your conf14:09
seiflotfyguys please14:10
alcabreraseiflotfy: are you running off of the latest master? sqlite has been removed as a backend.14:10
seiflotfyi am14:10
alcabreraWe now have sqlalchemy to provide that14:10
seiflotfyso the conf is broken then14:10
alcabreraI suspect that is the case14:10
seiflotfyok can someone show me how the conf should look like?14:10
alcabrerasure thing14:10
flwangseiflotfy: I think flaper87 has updated the sample conf14:10
seiflotfyi am using it and its not working14:11
flwangseiflotfy: ok, would you mind post yours? because it works for me, and let me share mine14:11
alcabreralemme help out on that14:11
seiflotfylast update was 15 days ago14:11
seiflotfyits http://paste.openstack.org/show/72180/14:12
alcabrerastorage = sqlite -> storage = sqlalchemy14:14
flwangseiflotfy: oh my gosh14:14
alcabrerathat should do it14:14
flwangseiflotfy: you're using mongo and its uri is not correct, IMHO14:14
flwangalcabrera: and seems there is another issue14:15
seiflotfythat is the default14:15
flwangalcabrera: I asked above, seems there is no check if the [drivers] storage = sqlite  NOT compatiable with [drivers:storage:sqlalchemy], did i miss something?14:15
flwangseiflotfy: you should use your correct storage backend14:15
alcabreraI'd expect storage = sqlite to trigger a Stevedore exception, flwang14:16
alcabrerasince sqlite was removed from setup.cfg:entry_points14:16
flwangseiflotfy: i'm saying, if you want to use mongodb as the backend, you should make sure it's installed on your env and update the uri in conf with the correct one14:16
seiflotfyno i dont want mongodb14:17
seiflotfybut this is the conf i got14:17
flwangalcabrera: nope, let me reword my concern14:17
alcabreraflwang, seiflotfy: I've got this one confirmed working as of now: http://paste.openstack.org/show/72183/14:18
alcabreraflwang: yes, thanks! :)14:18
* alcabrera listens14:18
flwangalcabrera: my conf works great, I'm just raising an concern14:18
flwangalcabrera: what will happend if I use [drivers] storage = mongo  with [drivers:storage:sqlalchemy]14:19
alcabrerawhat'll happen there, then14:19
flwangor the opposite14:19
*** ykaplan has joined #openstack-marconi14:19
alcabrerais that the mongo backend will be used14:19
alcabrerathe sqlalchemy options wqill be ignored14:20
flwangalcabrera: i don't know, haha, just think alound14:20
alcabreraand if no mongo options are specified, then the defaults are used14:20
alcabrerasame for the reverse case14:20
alcabrera'storage = <backend>' takes precedence14:20
alcabreraand determines the behavior from there on14:20
flwangalcabrera: so is it necessary to add a pre check?14:20
flwangalcabrera: ok14:21
alcabreraI don't think so. :)14:21
flwangalcabrera: I will dig14:21
alcabrerasomething I'd like to see14:21
alcabrerais INFO output14:21
alcabreraat start up14:21
alcabreraindicating what storage/transport driver was used14:21
flwangalcabrera: yep, it would be better14:22
alcabreraanother thought - sqlalchemy operating with the sqlite backend should be the default driver in marconi.conf-sample so that anyone can download marconi and start playing with it immediately14:22
flwangalcabrera: I love this suggestion14:23
alcabreracool. :)14:23
flwangalcabrera: it will be more friendly for the new comer14:23
flaper87kinda back14:24
flaper87alcabrera: goood morning!14:24
alcabreraflaper87: yo!14:24
flaper87seiflotfy: did you manage to fix that bug ?14:25
flwangalcabrera: I will file a wishlist bug to track that, thoughts?14:25
flaper87alcabrera: 'sup buddy?14:25
* flaper87 announces https://plus.google.com/events/ckjhm8rggnqvrnspftna845kubk14:25
flwangflaper87: thanks for the sharing14:26
flwangseiflotfy: did you try alcabrera's conf or mine?14:26
seiflotfywhere is alcabrera s?14:27
flaper87seiflotfy: what's the error? what's the problem your seeing ?14:27
flaper87and also, why do you need a config file?14:27
alcabreraseiflotfy: http://paste.openstack.org/show/72183/14:27
flaper87you shouldn't need one for that test, it's a unittest14:27
flaper87(i mean, for the one you were fixing)14:27
alcabreraflaper87: nice - re: G+ page14:27
flaper87alcabrera: that's the link to a Hangout I'll have on the 27th14:28
flwangalcabrera: btw, I think [drivers] storage = sqlite  will work fine, because flaper87 has considered the backward compatibility, IIRC based on the code review14:29
*** mpanetta has joined #openstack-marconi14:29
alcabreraflwang: you're right! I verified just a moment ago that sqlite points to sqlalchemy in setup.cfg14:29
* flaper87 gives himself a pop-tart for that14:29
flwangflaper87: let's give flaper87 a cookie :)14:29
flwangflaper87: haha14:30
alcabreraheh. :)14:30
flwangflaper87: I assume your comments will be coming soon on my baby patch, is it? :)14:33
flaper87ah damn, I forgot to click on submit14:33
flaper87that's how easy it is to distract me14:34
mpanettaMorning all14:34
flwangmpanetta: morning14:34
flwangalcabrera: it would be nice if you can take a look at my baby https://review.openstack.org/#/c/76555/14:35
alcabrerampanetta: good morning!14:35
alcabreraflwang: certainly. I'll check it out soon. :)14:35
flwangalcabrera: cool, thanks14:35
*** ametts has joined #openstack-marconi14:36
flwangalcabrera: I think we need to discuss the v1.1 feature in today meeting given the FF14:37
alcabreraoh *right*! I completely forgot about today's meeting!14:38
alcabrerathanks for reminding me, flwang!14:38
alcabreraAlso, I agree.14:38
flaper87holy moly, there's a meeting today14:38
flaper87mpanetta: morning14:38
alcabreraI'm currently at the dentist with Jess, so depending on when they finish up, I may miss part of the meeting14:39
alcabreraflwang: I added API v1.1 as a discussion item to the meeting agenda14:40
flwangalcabrera: cool, thanks14:41
*** amitgandhi has joined #openstack-marconi14:43
alcabrerabalajiiyer: good morning!14:43
*** amitgandhi has quit IRC14:44
*** amitgandhi has joined #openstack-marconi14:44
flwangflaper87: when will you be available to discuss those comments on my /health bp?14:45
alcabreraseiflotfy: awesome! glad to hear it14:49
flaper87flwang: shoot14:49
*** haomaiwang has joined #openstack-marconi14:49
flaper87seiflotfy: w0000t14:49
seiflotfyok pushing a small fix here14:49
balajiiyeralcabrera: Morning.14:49
alcabrerabalajiiyer: do you think you'll be up for discussing pecan evaluation at today's meeting?14:50
flwangflaper87: I collected the KPI from those ops guys, and they prefer to return the active shards number. but seems you have some concerns on that, is it?14:51
seiflotfy2014-03-04 15:51:28.977 10172 ERROR marconi.common.pipeline [-] Method does_not_exist not found in any of the registered stages   appears14:51
seiflotfyhow big of an issue is this14:51
balajiiyeralcabrera: yeah, I can talk about how Im doing the performance numbers. tsung + ab. have to do some due diligence before publishing the numbers.14:52
alcabrerabalajiiyer: cool, and of course. :)14:52
alcabreraI ask since pecan evaluation is on the agenda for today's meeting14:52
balajiiyeralcabrera: ok, I will be there14:53
alcabreraw00t, awesome14:54
*** malini_afk is now known as malini14:56
openstackgerritSeif Lotfy proposed a change to openstack/marconi: Fix assertion for Testutils to check on sqlite://:memory:  https://review.openstack.org/7790114:57
*** saikrishna_ has joined #openstack-marconi14:58
flaper87seiflotfy: that's not an issue14:58
flaper87I mean, IIRC, it's expected in those tests14:58
flaper87flwang: well, the number of shards is useful *from* the shard catalog14:59
seiflotfyyeah i noticed tests worked in both cases14:59
flaper87it depends on what shards we're talking about14:59
seiflotfywith or without the :14:59
*** kgriffs_afk is now known as kgriffs14:59
alcabrerameeting in 30 seconds14:59
alcabrerakgriffs: you made it!14:59
flaper87if we're talking about mongodb shards, we can't assume mongodb is sharded14:59
kgriffsice storm last night14:59
kgriffsmy car was totally encased in ice15:00
flaper87if we're talking about marconi's then it's necessary to get the shards from the shard catalog15:00
kgriffsfun times15:00
kgriffsmeeting time?15:00
flaper87kgriffs: fuck yeah15:00
kgriffsrock on15:00
flwangflaper87: got it, let's meeting first15:00
seiflotfy[13:40:12]  <@flaper87> seiflotfy: just replace it here: tests/unit/common/storage/test_utils.py15:00
seiflotfy[13:40:22]  <@flaper87> I mean, s/memory/:memory:/15:00
flaper87seiflotfy: what's up with that?15:02
seiflotfythis is the commit i did15:02
flaper87did I say something stupid?15:02
flaper87ah ok15:02
*** cpallares has joined #openstack-marconi15:03
seiflotfyflaper87 now fixing some performance issues15:09
*** alcabrera has quit IRC15:10
*** haomaiw__ has joined #openstack-marconi15:12
*** jergerber has joined #openstack-marconi15:12
seiflotfyflaper87 flwang i need help again with the test system15:13
flaper87seiflotfy: what performance issues?15:14
flaper87seiflotfy: shoot15:14
seiflotfythe nested functions15:14
seiflotfybut this is nto the issue15:14
seiflotfymy issue is that  somehow i change code and run the tests and they pass even if the code i write makes 0 sense15:15
*** haomaiwang has quit IRC15:16
seiflotfyflaper87 flwang15:17
flwangseiflotfy: yep, will review it, it's on my list now15:18
seiflotfywoops wrong link15:18
seiflotfyflwang flaper8715:18
seiflotfythis is what i get15:18
seiflotfywhatever stuff i change in my code it still passes15:18
seiflotfyhow come15:18
seiflotfye.g: in marconi/queues/storage/utils.py i wrote15:19
seiflotfy    for item in iterable:15:19
seiflotfy        yield Look15:19
seiflotfythere is no Look anywhere15:19
seiflotfyso i expect the tests to not pass15:19
seiflotfythey do15:19
seiflotfyflwang ^15:20
flaper87seiflotfy: in your latest paste, all tests passed15:20
seiflotfywhich shouldnt15:20
seiflotfyok again15:20
seiflotfyi am fixing a new issue15:20
seiflotfyproblem is i dont think the unit test run properly here15:21
seiflotfymarconi/queues/storage/utils.py line 10915:22
seiflotfydo a "yield SoFresh" instead of "yield Keyed(item)" and run the tests15:23
*** chandan_kumar has joined #openstack-marconi15:24
*** alcabrera has joined #openstack-marconi15:27
*** Kuwagata has joined #openstack-marconi15:29
*** ayoung-ZZzZzzZz_ is now known as ayoung15:30
*** jmckind has joined #openstack-marconi15:37
*** nidhi has joined #openstack-marconi15:49
*** nidhi has quit IRC15:53
flwangkgriffs: I think we need to figure out what's gap now to land all the v1.1 patches, make sense?15:58
kgriffsok, sure15:58
kgriffsso, we have /health to finish up - some comments there from flaper87 to address15:59
kgriffsalcabrera: need your eyes on this pls?! https://review.openstack.org/#/c/76093/15:59
kgriffsflwang: I started work on requiring x-project-id header15:59
kgriffsand removed my name from several of the other v1.1 patches15:59
kgriffsso you or alcabrera can pick those up if you like15:59
jay-atlHey Guys, so the consensus on the #openstack-dev channel seemed to be that the base url for Marconi in the service catalog should NOT have the version in the URL. But that the client should add the /v1 to the url... Does that sound right? Currently the client is not adding the version to the supplied url. So I'd have to fix the client..16:00
flwangkgriffs: I would like to land in the /health with less KPI and add more with bug tracking or even in Juno16:00
flaper87jay-atl: I think there was kind of a bug there16:01
flaper87I *think*16:01
flaper87I remember digging into that at some point16:01
kgriffsflwang: +116:01
kgriffslet's do the bare minimum16:01
flwangflaper87: any concern?16:01
flaper87flwang: sounds good16:01
flwangrock on16:02
alcabrerakgriffs: +2/approved bulk delete. the version_changed doc directive is sweet!16:02
kgriffsalcabrera: Can you respond to the [marconi] email on the ML?16:02
alcabrerameeting minutes updated16:02
flwangkgriffs: I will see what I can help on the other v1.1 bp after I submit a new ps for /health16:02
* alcabrera missed the marconi email16:02
alcabrerasure, I can reply. I just need to find it.16:02
kgriffsalcabrera: should be the latest one from openstack-dev16:02
alcabreraah, I see it16:03
kgriffs"when we compare tools we should be careful. Compare for understanding, not for judgement.  "16:05
*** saikrishna_ has quit IRC16:05
kgriffs(quote by Henrik Kniberg)16:06
alcabreraflwang: very +1 - I'd love to have you join us as a core-reviewer. :)16:06
kgriffsok folks, closing the poll16:06
flwangalcabrera: thank you, I will contribute more :)16:08
flwangthanks kgriffs, alcabrera, flaper87 and all you guys voting16:08
flaper87flwang: you better do16:08
kgriffsflwang: you now have +2 super powers16:09
cpallaresflwang: yay :)16:09
kgriffs"with great power comes great responsibility" and all that...16:09
flwangflaper87: I will be the top3 reviewers in Juno for Marconi16:09
maliniI am doing some lobbying on #openstack-qa to get my band-aid approved..Feel free to show your support16:09
flwangkgriffs: yep, I totally understand. but I like the challenge :)16:10
flaper87kgriffs: that's a nice way to put it16:11
flwangflaper87: so back to the /health, I would suggest to postpone the server status stuff, any concern?16:12
flaper87flwang: meaning, just doing the storage drivers part ?16:12
flwangflaper87: I would like to know kgriff and alcabrera's comments16:13
* alcabrera was going to run off for a bit, but waits to give comments16:13
flwangexcept the server cpu util and memory util, any info we'd like to deliver from the storage backend?16:14
alcabreraflwang: you mean the KPIs?16:14
flwangalcabrera: yep16:14
alcabreraI'm fine with  those being put off. :)16:14
flaper87sounds good16:14
alcabrerafor me, as long as /ping is a shallow check and /health allows for a deep check, I'm happy16:14
*** saikrishna_ has joined #openstack-marconi16:14
flwangalcabrera: yep, we can add more KPI in the future16:15
*** jamieh has quit IRC16:15
alcabrerafor sure16:15
alcabreraI'm going to run off for a bit16:15
alcabreraheading into the office. :D16:15
alcabrerasee you all in 30-ish16:15
flwangalcabrera: see you later16:15
mpanettabe safe16:15
mpanettasee ya when ya get here16:15
*** alcabrera has quit IRC16:15
maliniI am getting ready to be in a mood to rant :-@16:16
*** haomaiw__ has quit IRC16:16
*** haomaiwang has joined #openstack-marconi16:16
malinikgriffs, flaper87: https://review.openstack.org/#/c/7783216:21
*** haomaiwang has quit IRC16:21
maliniSee the last review comment16:21
flaper87I'm really sick of those comments16:22
flaper87for gods sake16:22
mpanettaOh hell...16:23
mpanettaAre we plagued by highschoolers?16:23
maliniI thought I was working in a 'community' !16:23
mpanettaHighschool community? :P16:24
maliniIts not as if I am corrupting a pristine code base with a redirect16:24
flaper87not highschoolers at all and it's not his fault. All those guys are really cool it's just that this stuff about us using falcon is being used as an excuse to block some patches that are critical16:25
mpanettaBut why?16:25
maliniwhat has a lo redirect got to do with falcon?16:25
mpanettaThat seems so... I donno... Childish?16:25
malini& blocking critical patches coz of the falcon thing, is not a professional response16:25
kgriffsoh boy16:25
flaper87malini: exactly16:25
kgriffsI am about to write something...severe.16:25
flaper87kgriffs: I'll let you tackle it this time16:26
flaper87I'm like seriously frustrated now16:26
maliniI worked with two other devstack core reviewers to work on this16:26
malini& nobody has any clue16:27
maliniso its not that we didnt try hard enough16:27
mpanettaI still think it is all the odd redirects that are going on...16:27
*** vkmc has quit IRC16:27
mpanettaIn the devstack code.16:27
malinimpanetta: but we need to prove its a devstack issue.16:28
maliniThe thing is we dont where tht is happening16:28
mpanettaThat should be easy16:28
mpanettareplace marconi-server with a shell script that does nothing but loops and prints to stdout16:28
mpanettaIf *that* crashes then it has to be a devstack issue16:29
malinimpanetta: can you do that?16:29
mpanettaWriting to stdout should not cause a crash16:29
kgriffsum, ok16:29
kgriffsfire me16:29
kgriffskill me16:29
maliniI am too sick & tired to try anything now16:29
* kgriffs commented16:29
mpanettaLet me send you a gist that you can pull down16:29
maliniI need to take a break now..I am brain dead to work on anything16:30
malinibe back in 30 min16:30
mpanettaEr, crap I have to wait until after this meeting that starts in a minute...16:30
*** saikrishna_ has quit IRC16:30
flwangflaper87: kgriffs: please revisit my baby when you're available https://review.openstack.org/#/c/76555/16:32
flwangkgriffs: seems I still don't have the superman power, did you add me on the gerrit?16:33
kgriffsflwang: let me check16:33
flwangkgriffs: I'd like to be a superman :)16:34
kgriffstry now16:34
flwangkgriffs: i'm a superman now, guru16:34
kgriffsrock on16:34
kgriffssorry about that16:34
flwangkgriffs: np16:34
kgriffsflaper87: check me on this logic16:34
* flaper87 checks kgriffs circuits16:35
*** whenry has joined #openstack-marconi16:35
kgriffsstand by ... let me check something16:35
flaper87malini: is it possible to have a custom localrc in the gate?16:35
flaper87I don't remember anymore16:35
flaper87or at least setting custom env variables16:36
flaper87malini: in other words, I want to set this to sqlalchemy https://github.com/openstack-dev/devstack/blob/master/lib/marconi#L44 :D16:36
flaper87but *just* in the gate job16:36
mpanettaflaper87: I think she went away to blow off some steam.16:37
flaper87mpanetta: hahahahahaha16:38
mpanettaflaper87: Are you next? :P16:38
* flaper87 is THE saint16:38
seiflotfyflaper87 why are 290 tests on my side being skipped?16:40
seiflotfyPASSED (id=18, skips=290)16:41
flwangseiflotfy: IIRC, it's db specific16:42
*** saikrishna_ has joined #openstack-marconi16:42
flaper87seiflotfy: because mongodb tests are not enabled by default16:45
seiflotfyhmmmm can i enable them too16:45
seiflotfyor do i need to change my db?16:45
flaper87seiflotfy: export MARCONI_TEST_MONGODB=116:46
seiflotfythe problem i have right now i cant reliably test stuff16:46
kgriffsok, flaper87: check me on this16:48
kgriffshas self._wsgierrors = env['wsgi.errors']16:48
kgriffsthat is used to implement log_error16:48
kgriffsnow, under wsgiref, that probably means writing to the console16:49
kgriffsok, yeah16:49
kgriffsBUT, malini proved by redirecting output to a file that it is zero bytes output16:49
kgriffsand so, log_error could not be the culprit16:49
*** alcabrera has joined #openstack-marconi16:49
kgriffsI will put a breakpoint in there just in case16:50
seiflotfyflaper87 basically anything i change in marconi/queues/storage/utils.py16:50
kgriffsbut if that IS the problem, saying this is a "bug" that is Falcon's fault is like saying this is the WSGI spec16:50
seiflotfyanything i change in marconi/queues/storage/utils.py even it does not make sense and i yield non existing classes the tests pass16:50
kgriffsseiflotfy: looking16:53
flaper87kgriffs: mhhh, that looks correct, I'm not sure what wsgiref does, though16:53
mpanettaOk this is weird16:53
mpanettaI replaced marconi-server with a short C program that prints a line every second and I still see the same output from the devstack scripts, and it kills my fox.16:54
seiflotfykgriffs http://paste.openstack.org/show/72246/ is me running through the tests16:55
seiflotfywhy is it trying keystone16:55
alcabreradevstack - notorious fox slayer16:55
mpanettayeah I called the program fox...  It just prints "A quick brown fox jumps over the moon" every second heh16:55
flwangflaper87: based on my understanding, the active shard number is the row count we stored in Shards table16:55
mpanettaThis looks like it isn't a marconi issue to me...16:55
mpanettaEither that or they require that we background and daemonize ourselves...16:56
mpanettaElse they will kill us16:56
alcabrerahmmm, I wonder who the person to talk to about devstack internals is...?16:56
flwangflaper87: so I got the number by accessing db for storge driver, but get it by shard catalog for sharding, pls correct me if there is anything wrong16:56
alcabreraI have to say, the devstack docs are beautiful: http://devstack.org/16:57
flaper87flwang: but you shoulnd't be getting the number of shards in the DataDriver16:58
flaper87that's what my comment was about (assuming you're talking about marconi shards )16:58
flaper87flwang: alcabrera can assist you better on the shard part16:58
flwangflaper87: ok, so double confirm, I will postpone the server status of mongo to the future, agree?16:59
flwangfor now, let's just do cpu_util, memory_util and active shard number16:59
mpanettaalcabrera: Wow they do have extensive docs...17:00
maliniRunning is a good anger management technique17:00
flwangalcabrera: may I get some help from you about the shard part?17:00
alcabreraflwang: sure thing17:00
mpanettamalini: Devstack killed my fox :(17:00
alcabreramalini: agreed. I danced (DDR'd) this morning to prepare for the day. :D17:01
malinimpanetta: ?17:01
flwangalcabrera: pls take a look before I ask some questions :)17:01
*** amitgandhi has quit IRC17:01
mpanettaIt behaves the same as marconi-server, prints one line and dies17:01
*** amitgandhi has joined #openstack-marconi17:01
mpanettaBut my fox program prints forever17:02
mpanettaSo devstack must be killing it17:02
maliniI found some weird comments y'day abt redirecting stdout17:02
mpanettaI donno what it means but maybe it will help.17:02
malinilet me find tht17:02
alcabreraweird, mpanetta.17:02
maliniI did chk with dtroyer a few times while hunting this one17:02
alcabreraflwang: where am I taking a look? Is there a patch out already> :)17:03
maliniBut he didnt mention anything abt the stdout stuff17:03
flwangalcabrera: yep :) https://review.openstack.org/#/c/76555/17:03
mpanettaWell I wrote my fox in C so it isn't a python thing...17:03
* alcabrera reads the review17:04
flwangalcabrera: and another baby https://www.dropbox.com/s/8xpojqy6ii3b5pc/mmexport1386998377660.jpg?m=17:05
mpanettaalcabrera: Yes quite weird17:05
alcabreracuuuute, flwang! <317:06
flwangalcabrera: 5 months :D17:06
alcabrera+2/Approved instantly on the other baby. :D17:06
cpallaresflwang: what a cute baby!17:06
mpanettaYes, but what does jenkins think? :P17:06
flwangalcabrera: haha, on my plan17:06
flwangmpanetta: good question, hehe17:07
kgriffsmalini: does any test succeed in devstack?17:07
kgriffsor does it fail on the first one?17:07
malinikgriffs: all of them fail with the same connection error17:10
maliniI am going to work with mpanetta to figure out why stdout irritates devstack17:11
kgriffsI just noticed that marconi-server -v is logging each request to my console17:12
kgriffseven though I have a file path configured17:12
kgriffsit isn't getting logged though falcon because I commented out Request.log_error17:13
kgriffsso I believe I was correct in saying this bug has NOTHING to do with Falcon vs. Pecan17:14
*** rwsu has joined #openstack-marconi17:14
kgriffsand even if it did, it would be stupid that devstack couldn't handle writing out to the WSGI error stream17:14
*** ykaplan has quit IRC17:15
* alcabrera starts reviewing flwang's patch for context17:16
flwangalcabrera: appreciate it :)17:16
flwangalcabrera: working hard on that17:17
alcabrerathanks! :D17:17
* flwang rock on17:18
malinikgriffs: mpanetta just proved its not a WSGI thing with his fox app17:19
maliniTempest is exactly that :D17:23
kgriffsmalini: thanks!17:23
maliniI am debugging the devstack script to find the point of failure17:24
alcabreraEverytime I start typing Tempest, I typo is as TempTest17:24
alcabreraFreudian slip, maybe17:24
malinithis is where it fails https://github.com/openstack-dev/devstack/blob/master/functions-common#L97117:26
* alcabrera continues reviewing flwang's patch - many comments being written!17:27
flwangalcabrera: waiting from Beijing, China, at 1:28AM :)17:28
*** reed has joined #openstack-marconi17:30
seiflotfyflaper87 any other way to run tests here17:31
alcabreraflwang: hahaha, get some sleep! ;)17:32
alcabrerathe review is done now17:32
flwangalcabrera: reading... I want to make it land in ;017:32
alcabreraI -1'd, primarily because 'active_shards' shouldn't be present all the time17:32
alcabreragotcha, flwang. :)17:32
alcabreragoing to go fetch a sprite zero. nobody mutate the world while I'm away. ;P17:34
*** alcabrera is now known as alcabrera|afk17:34
*** alcabrera|afk is now known as alcabrera17:35
flwangflaper87: any lucky around?17:36
kgriffsmalini: https://gist.github.com/anonymous/29619594145df6db5cf217:36
kgriffsI'm trying to figure out where that message is coming from...17:36
kgriffswsgiref itself?17:36
kgriffslooks like a vanilla web server loglline17:37
malinikgriffs: yes..& It writes to both stdout & the log_file17:38
kgriffsi would say "that's our problem" except when you do the redirect workaround, the result file is 0 bytes!17:41
maliniif its ok for devstack to kill an app with stdout, it is OUR problem17:45
*** jraim_ has joined #openstack-marconi17:45
openstackgerritKurt Griffiths proposed a change to openstack/marconi: feat(api): API v1.1 bulk delete change  https://review.openstack.org/7609317:46
kgriffsmalini: sorry, I meant that wsgiref printing out is "the problem"17:46
kgriffsnot a marconi vs. devstack thing17:46
kgriffsI found it curious that your workaround wrote 0 bytes17:47
alcabrerawsgiref.silent = True (sure would be nice, heh)17:47
kgriffsif wsgiref is writing out stuff17:47
maliniit wud have been good if people can work together to make things better, instead of you don't do stuff like me <sigh/>17:48
malinitht is the reason world has wars17:49
*** mwagner_lap has joined #openstack-marconi17:49
kgriffsthat is a BIG PROBLEM with a subset of OpenStack community17:49
kgriffsI'm tired of people saying "well, that's just the way OpenStack has always been"17:50
maliniI am going to end up a hippie with a CO-EXIST bumper sticker, after this17:50
alcabreraheh, malini. :D17:50
alcabreraget me one, too17:50
*** alcabrera is now known as alcabrera|afk17:51
*** chandan_kumar has quit IRC17:53
*** jraim has quit IRC17:53
*** jraim_ is now known as jraim17:53
kgriffsmalini: I don't see those request loglines in marconi.log17:53
kgriffsso I suspect it is wsgiref doing it17:53
kgriffs(on my box, at least17:53
*** balajiiyer has quit IRC17:55
*** alcabrera|afk is now known as alcabrera17:55
flwangis there any possibility the marconi.conf will be cached?17:55
flwangafter changed the storage driver from sqlalchemy to mongodb, seems it's still using the sqlite17:56
kgriffsflwang: after restarting the wsgi server, it should pick up the new one17:56
kgriffsmaybe it isn't finding your marconi.conf and defaulting to sqlalchemy?17:56
flwangkgriffs: I assume that as well17:57
kgriffsalcabrera: ping17:57
flwangbut I have one under ~/.marconi17:57
flwangkgriffs: when is the next checkpoint for v1.1 today?17:57
kgriffsI guess it would be in a few minutes17:58
kgriffsso, I was distracted by tempest thing17:58
kgriffsthis has been such a time sink17:58
kgriffsI just now started working on v1.1 things17:58
alcabrerakgriffs: pong17:59
flwangi think the overall shape of /health is ok based on the comments from flaper87and alcabrera17:59
kgriffsalcabrera: can your write a quick wsgiref "hello world" app to confirm it is logging and that isn't anything to do with MArconi?17:59
kgriffsi mean, logging to stdout17:59
alcabrerasure thing17:59
openstackgerritShawnyu proposed a change to openstack/marconi: Use sqlalchemy as the default storage driver, and update it in sample configuration file  https://review.openstack.org/7794917:59
flwangseems the biggest concerns is the KPI - active shards number, I'm think replace it with a simple one17:59
alcabrerahave it in a paste in 517:59
kgriffsflwang: ok, let's simplify or remove altogether18:00
alcabrerakgriffs: desired log level? DEBUG> INFO?18:00
kgriffsshouldn't matter18:00
kgriffswsgiref doesn't respect logging level18:01
flwangkgriffs: i think cpu_util and memory_util are both ok, you know, they are related to the marconi SERVER, and we're trying to add another KPI from the storage backend perspective18:01
kgriffsI just tested running simple_server with simple_server.demo_app and it is writing out the same request lines to console18:01
alcabrerakgriffs: http://paste.openstack.org/show/72261/18:02
*** balajiiyer has joined #openstack-marconi18:03
kgriffsactually, I was thinking to not use logging in there18:03
kgriffsbasically a noop18:03
flwangkgriffs: so I'm thinking can we just leverage the is_alive()'s result and return it as a part of the detailed /health, just think alound18:03
alcabreraI see18:03
alcabreraI'll simplify it18:03
kgriffsand see if when you run with simple_server.make_server if it still writes to console18:03
kgriffsflwang: does is_alive return anything now?18:04
alcabrerakgriffs: http://paste.openstack.org/show/72262/ (barebones)18:04
alcabrerayeah, I'll try that and report18:04
flwangtrue of false to indicate if the backend storage is alive or not18:04
flwangalcabrera: pls check me if i'm wrong18:04
kgriffsflwang: ok, we could just report that as a field "storage_reachable: True/False" or something18:05
kgriffsat this point, I'd just say keep it super simple and we can add more KPIs later18:05
flwangalcabrera: any comments about replacing the active shards with storage_reachable ?18:06
kgriffswe just want to demonstrate pulling one KPI from storage18:06
flwangkgriffs: yep18:06
kgriffsflwang: we could also count total number of messages per shard or something18:06
kgriffsthat should be easy to implement18:06
kgriffs(in each driver)18:06
kgriffsbut, I think if we can just use the alive call that is good enough for now18:07
flwangkgriffs: yep, I discussed that with alcabrera recently, but seems it's a little bit difficult18:07
kgriffslet's KISS18:07
flwangkgriffs: ok, will submit a ps soon18:07
alcabrerakgriffs: it logged nothing on start up, but....18:07
alcabrera127.0.0.1 - - [04/Mar/2014 13:07:31] "GET / HTTP/1.1" 200 018:07
kgriffsalcabrera: ok18:07
kgriffsthen that is wsgiref18:07
kgriffsmalini: ^^^ this could be the culprit18:07
kgriffsalthough why other projects don't have the same issue, is beyond me18:08
kgriffsor they have a hack to work around it18:08
kgriffsalcabrera: can you find a way to supress those messages?18:08
alcabrerafor ref: http://paste.openstack.org/show/72263/18:08
alcabrerasure thing18:08
kgriffsthen malini can try your solution to see if it helps anything18:08
kgriffsalcabrera: thanks!18:08
malinisure thing..18:08
maliniI hope this fixes it18:08
kgriffsok, I am going to be back in 5 mins and then dive into v1.1 stuff18:08
*** rossk has joined #openstack-marconi18:10
alcabrerakgriffs, malini: solved - http://paste.openstack.org/show/72269/18:14
*** fvollero is now known as fvollero|afk18:14
alcabreraI hope. It doesn't log for me. :D18:14
alcabrerahttp get localhost:8001  # was very quiet18:14
kgriffsok, malini can you try that?18:17
kgriffsif it works, we could have log_message write to LOG and see if it still works18:18
kgriffs(it should as long as a path is set in marconi.conf for the log file)18:18
kgriffskk, brb18:18
kgriffsI need to cool down18:18
openstackgerritFei Long Wang proposed a change to openstack/marconi: feat(v1.1): Implement new /health endpoint for v1.1  https://review.openstack.org/7655518:18
flwangflaper87: alcabrera: please help review  https://review.openstack.org/76555 :)18:21
alcabreramalini: marconi/marconi/queues/transport/wsgi/driver.py:listen18:25
alcabreraapply the wsgiref fix there18:25
alcabrerathen we can all18:25
flwangalcabrera: flaper87: i'm removing the active shards with storage reachable which can leverage current is_alive to get a basic KPI from the storage backend POV18:26
alcabreramuch ☮, such ♥18:26
*** notmyname has left #openstack-marconi18:26
flwangplease let me know if you have any concern/comment about this18:26
alcabreraflwang: I'll look at the health patch again soon. :)18:27
flwangalcabrera: appreciation from China :)18:27
*** fvollero|afk is now known as fvollero18:28
*** balajiiyer has quit IRC18:33
mpanettaHere, to make every ones day better: http://i.imgur.com/jsmh5t0.gif :P18:35
flwanghaha, I guess the foot is smelly :)18:38
* alcabrera catches up18:44
malinikgriffs: still no luck :(18:45
alcabreraso far, redirecting to /dev/null feels right to me. :/18:46
alcabreraI'm going to go read some devstack18:46
alcabreraand cry a little18:46
malinijoin the club alcabrera !18:46
flwangi'm going to take some sleep18:46
flwangkgriffs: when is the next checkpoint?18:47
openstackgerritCaleb Groom proposed a change to openstack/python-marconiclient: Add docstrings to modules to reduce pylint score  https://review.openstack.org/7795818:47
flwangalcabrera: pls feel free drop you comments on my baby patch, I will be back after 5, 6 hours if I can wake up :D18:48
alcabreraflwang: good night! :)18:50
flwangalcabrera: have a nice day, guys18:51
openstackgerritFei Long Wang proposed a change to openstack/marconi: feat(v1.1): Implement new /health endpoint for v1.1  https://review.openstack.org/7655518:51
*** balajiiyer has joined #openstack-marconi18:53
malinialcabrera, kgriffs: looks familiar https://github.com/openstack-dev/devstack/blob/master/lib/trove#L205 ?18:55
kgriffsflwang: thanks man18:55
*** karandash8 has joined #openstack-marconi18:55
*** balajiiyer1 has joined #openstack-marconi18:55
kgriffssorry, too many distractions18:55
*** cpallares has quit IRC18:55
kgriffsmalini, alcabrera: any luck?18:56
malinikgriffs: not with the WSGI change18:56
malinibut trove seems to be doing some weird redirections too18:56
mpanettaThe change did not help? :(18:57
malinimpanetta: no18:57
maliniwonder if trove folks ran into something similar18:57
*** jmckind has quit IRC18:57
mpanettaThis is so very odd18:57
*** balajiiyer has quit IRC18:58
maliniI see alot of re-direct to /dev/null in other projects18:58
malinithere shud be something going on18:58
mpanettaThen why can't we do that?18:58
maliniwe need to figure out why they did it18:59
mpanettaCan we ask them?18:59
malinilet me try doing a blame to figure out whom to ask19:00
*** SnowDust has joined #openstack-marconi19:01
maliniMichael Basnight wrote tht part19:01
malinikgriffs: he is a RAX guy, rt?19:02
alcabreramalini: yes19:02
alcabreraI just checked. :)19:02
maliniI hope he is in Austin & sits next to kgriffs ;)19:02
SnowDustmalini is who ?19:02
maliniMichale Basnight19:03
maliniSnowDust: ^19:03
SnowDustmichael basnight ? whozz hub_cap ?19:03
malinihmmm..dont know19:03
SnowDustROFL ..19:04
maliniI think my humor sense is kinda dead today19:04
*** amitgandhi has quit IRC19:04
SnowDustmalini is an indian word btw ..19:05
malininot surprising,considering I am from India19:05
SnowDustMichael Basnight from india ?19:05
SnowDustits not shocking anymore ..19:06
SnowDustmay be rackspace shifted to lower parel  or nehru place .. who knows19:06
malinimy brain is already messed up..I think I can do without more help :)19:06
*** sballe has quit IRC19:07
*** rossk has quit IRC19:08
* alcabrera is thoroughly confused19:08
mpanettaJoin the club heh19:08
* mpanetta creates a Club for the onfused19:08
mpanettaand unable to type, apparently19:08
*** amitgandhi has joined #openstack-marconi19:09
kgriffsok folks, sorry, but I am getting pulled into the Oslo mailstrom19:09
alcabrerakgriffs: good luck!19:10
kgriffsso, I am going to make the call and push v1.1 to Juno19:10
kgriffsanyone vehemently disagree?19:11
openstackgerritA change was merged to openstack/marconi: fix(mongodb): Limit kwarg must be an int  https://review.openstack.org/7667819:11
alcabreraI don't disagree, kgriffs19:11
alcabreraI think it's the right call19:12
alcabreragiven we need to discuss some of the coming fixes a bit more19:12
maliniI might have found something19:22
maliniif we do the same thing as trove, it works too19:22
maliniuse '--debug 2>&1' instead of '>/dev/null'19:22
mpanettaSo redirecting stderr to stdout worked?19:23
mpanettaOh you don't need the --debug19:23
mpanettaJust the 2>&1, sorry19:23
maliniok let me try without tht19:23
mpanetta--debug is a flag for whatever they are running19:23
malinibut waht does tht mean?19:23
maliniI mean 2>&1 ?19:23
mpanettaThat means redirect stderr to stdout19:24
mpanettaso all output (error or not) goes to stdout.19:24
mpanettaAKA FD 119:24
malinihmm..but tht means our original theory is wrong abt stdout?19:24
alcabreradevstack seems contradictory19:25
malinior it had to be in some format for functions-common to understand?19:25
mpanettathe 2>&1 thing is a bashism19:25
alcabreraI vaguely remember a higher-level command redirecting stderr to /dev/null19:25
maliniI hope I am not imagining this to be working :D19:25
*** jamieh has joined #openstack-marconi19:26
*** alcabrera is now known as alcabrera|afk19:27
malinimpanetta: it works without the debug too19:28
mpanettaGood, that is what I would expect :)19:28
maliniIt feels weird going back & saying 'they did it, so we can too'19:28
mpanettaSomeone somewhere is screwing up the normal file descriptors I think.19:29
maliniBUT I am going to go back & ask19:30
maliniwish me luck19:30
*** alcabrera|afk is now known as alcabrera19:31
alcabreramalini: good luck!19:31
*** alcabrera is now known as alcabrera|afk19:33
*** mpanetta has quit IRC19:47
*** mpanetta_ has joined #openstack-marconi19:47
*** cpallares has joined #openstack-marconi19:51
*** mpanetta_ is now known as mpanetta19:51
*** alcabrera|afk is now known as alcabrera19:54
maliniI need to leave now..I am still waiting to hear back from Sean Dague on if the other alternative is acceptable for https://review.openstack.org/#/c/77832/19:57
maliniIf I hear back from, I'll submit a new patchset19:57
*** malini is now known as malini_afk19:57
mpanettaI hope you get something back today20:02
openstackgerritShawnyu proposed a change to openstack/marconi: Use sqlalchemy as the default storage driver.  https://review.openstack.org/7798520:03
openstackgerritShawnyu proposed a change to openstack/marconi: sqlalchemy: make default storage driver  https://review.openstack.org/7798620:06
kgriffsflaper87: wrt bugs and the feature freeze20:07
kgriffsactually, let me go find that on the wiki20:07
alcabreraraxuanyu: Let's get together tomorrow. You don't need to submit a new patch for each fix. The way to get around that is that each time you make a fix to a given patch, you need to rebase against master.20:08
*** amitgandhi has quit IRC20:09
kgriffsalcabrera, flaper87: the way I read this page is that we should triage remaining bugs and only leave the ones for i-3 that are "milestone-critical"20:11
alcabreraRelease Candidates dance - nice20:12
openstackgerritShawnyu proposed a change to openstack/marconi: sqlalchemy: make default storage driver  https://review.openstack.org/7798820:13
alcabreraI agree with your interpretation, kgriffs20:13
alcabreraso, we need to triage at some point20:13
kgriffsactually, we need to do it now20:14
kgriffsrelease mtg in 45 minutes20:14
kgriffsflaper87: ^^^20:15
kgriffslooks like it is just you and me20:27
kgriffslet's get to it20:27
*** SnowDust has quit IRC20:29
kgriffsGet a set of messages returns wrong hrefs20:32
kgriffshas anyone reported this bug in the wild?20:33
alcabreraI haven't heard about it20:33
kgriffsthis report is for the gate20:33
kgriffsok, sounds like we can move this one20:33
alcabrerayes - non-critical20:33
kgriffsadd functional tests for sharding20:34
alcabreranon-critical in one sense20:36
alcabrerait doesn't affect i3 features20:36
alcabrerawe need soon after20:36
alcabreraso a level below critical20:36
alcabrerakgriffs: thoughts? :)20:36
*** saikrishna_ has quit IRC20:40
kgriffssorry, someone mentioned pecan/falcon in #openstack-meeting20:42
kgriffsalcabrera:  OK, we can keep as high20:42
kgriffsBut move to Juno?20:42
alcabrerano worries20:42
alcabreraand yes20:42
kgriffsnext: functions defined inline20:43
kgriffsthought I saw someone was working on that?20:43
alcabreraperf, right?20:45
alcabreraI don't remember anyone mentioning that for several months20:46
kgriffsah, I think Seif recently picked it up to work on it20:47
kgriffsI don't see a patch for it, do you?20:47
*** malini_afk is now known as malini20:47
alcabrerain any case20:48
alcabreraI'm calling it low20:48
alcabrerakgriffs: ^20:48
kgriffsLet's move this to Juno20:51
kgriffswrong clipboard20:51
alcabreramigration from ddt to testscenarios - low20:51
*** amitgandhi has joined #openstack-marconi20:52
alcabrerakgriffs: looking at this list, all the low bugs are still low, imho20:53
*** amit has joined #openstack-marconi20:53
alcabreraI'll tackle [bug/tests] ShardTests.test_listing_simple assumes alphabetical sorting order during the Summer, while working with GSoC students20:54
*** amit is now known as Guest9933620:54
kgriffsOK, can you remove icehouse and the milestone from each of the Low's for me?20:56
alcabrerasure thing20:56
*** amitgandhi has quit IRC20:57
*** jmckind has joined #openstack-marconi20:58
kgriffsalcabrera: status on this? https://bugs.launchpad.net/marconi/+bug/125861821:00
alcabrerakgriffs: there, that took care of it21:00
alcabrerastatus unknown, tbh. :/21:00
alcabreraI stepped away from that so long ago I'd need to check21:00
kgriffsok. I thought it was fixed, but I could be wrong21:00
kgriffsgo ahead and check21:01
alcabrerakgriffs: looks fixed: https://github.com/openstack/marconi/blob/master/marconi/queues/storage/sharding.py#L42521:03
*** Guest99336 has quit IRC21:03
*** amitgandhi has joined #openstack-marconi21:04
alcabrerait even got the unset_many update from oslo.cache v121:04
kgriffscan you update the bug accordingly?21:04
alcabrerasure thing21:04
kgriffsdrop a link to the fix or something in the whiteboard21:04
alcabrera"The fix to this was not tracked. It's been fixed at some point in the past month."21:06
*** jamieh has quit IRC21:09
kgriffswhatever works, man!21:10
kgriffsok, so looks like we have one bug pending, the doc patch pending, and the tempest patch pending21:10
alcabreravery close21:12
kgriffslet me check on my bug21:12
kgriffsas far as docs and tempest go...21:13
balajiiyer1kgriffs: do we have some extra time for doc patch?21:13
kgriffsbalajiiyer1: I don't know21:13
kgriffsI think they said it may be OK to let us graduate if the docs are ready within a week or two of i-321:13
kgriffsgraduating is sort of a strange thing21:14
kgriffsI think that if we show the patch is pending, that should be proof enough for them21:15
balajiiyer1yeah, we are very close with docs though. I will talk to Catherine and see whether she can submit a patch today.21:15
kgriffsthat would be great21:15
*** rossk has joined #openstack-marconi21:15
*** flaper87 is now known as flaper87|afk21:20
maliniI uploaded a new patchset https://review.openstack.org/#/c/77832/21:24
malinithis replaces the >/dev/null to do something similar to what trove folks did21:24
alcabreraI just saw that21:25
alcabrerathanks, malini!21:25
maliniI dont know if it is any better than the prev approach21:25
maliniBut has a precedent21:25
mpanettamalini: Awesome!  They better accept it :P21:25
maliniI dont know abt tht :)21:25
maliniI need to log off now21:26
alcabrerame, too21:26
maliniwill chk back later on the patch21:26
alcabreraGood night, all!21:26
maliniGood night!!21:26
*** malini has left #openstack-marconi21:26
mpanettaSee you guys lay=ter21:26
*** alcabrera has quit IRC21:27
*** Kuwagata has left #openstack-marconi21:41
*** AndChat|437241 has joined #openstack-marconi21:53
*** AndChat|437241 has quit IRC21:54
*** rossk has quit IRC21:57
*** rossk has joined #openstack-marconi21:58
openstackgerritCatherine Richardson proposed a change to openstack/marconi: adds docs directory with dev guide  https://review.openstack.org/7200122:02
*** tedross has quit IRC22:02
balajiiyer1kgriffs: Doc patch has landed from Catherine. It lgtm. can you take a look?22:18
kgriffsgive me just a minute22:18
*** balajiiyer1 has quit IRC22:25
openstackgerritKurt Griffiths proposed a change to openstack/marconi: fix(MongoDB): Driver does not retry on AutoReconnect errors  https://review.openstack.org/7516522:42
kgriffsbe back later folks. We will need to land the remaining patches by morning (EU time)22:50
*** mpanetta has quit IRC22:58
*** kgriffs is now known as kgriffs_afk23:00
*** amitgandhi has quit IRC23:16
*** jergerber has quit IRC23:25
*** cpallares has quit IRC23:43
*** mwagner_lap has quit IRC23:44

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