Monday, 2014-02-17

« Sunday, 2014-02-16 Index Tuesday, 2014-02-18 »
*** oz_akan_ has joined #openstack-marconi00:28
*** oz_akan_ has quit IRC00:32
*** flaper87 is now known as flaper87|afk00:36
*** nosnos has joined #openstack-marconi01:05
*** oz_akan_ has joined #openstack-marconi01:53
*** malini is now known as malini_afk02:57
*** oz_akan_ has quit IRC03:29
*** haomaiwang has quit IRC03:32
*** haomaiwa_ has joined #openstack-marconi03:33
*** haomaiw__ has joined #openstack-marconi03:35
*** haomaiwa_ has quit IRC03:38
*** flwang has joined #openstack-marconi03:39
*** ayoung has joined #openstack-marconi03:55
*** oz_akan_ has joined #openstack-marconi04:20
*** oz_akan_ has quit IRC04:25
*** rektide has quit IRC04:32
*** rektide has joined #openstack-marconi04:32
*** chandan_kumar has joined #openstack-marconi05:02
*** ayoung has quit IRC05:12
*** whenry has quit IRC06:10
*** saju_m has joined #openstack-marconi06:15
*** whenry has joined #openstack-marconi06:21
*** balajiiyer has joined #openstack-marconi06:36
*** nosnos has quit IRC06:42
*** nosnos has joined #openstack-marconi06:52
*** balajiiyer has quit IRC07:06
*** flaper87|afk is now known as flaper8708:00
*** ykaplan has joined #openstack-marconi08:20
*** saju_m has quit IRC08:31
openstackgerritKai Zhang proposed a change to openstack/marconi: Fix two small typo in marconi/queues/storage/sharding.py Replace 'preferabaly' with 'preferably' and 'fron' with 'from'  https://review.openstack.org/7396908:33
openstackgerritKai Zhang proposed a change to openstack/marconi: Fix two small typos in marconi/queues/storage/sharding.py Replace 'preferabaly' with 'preferably' and 'fron' with 'from'  https://review.openstack.org/7396908:35
*** saju_m has joined #openstack-marconi08:58
*** haomaiw__ has quit IRC09:02
*** haomaiwa_ has joined #openstack-marconi09:03
*** chandan_kumar has quit IRC09:10
*** flwang has quit IRC09:15
*** chandan_kumar has joined #openstack-marconi09:20
*** ykaplan has quit IRC10:04
*** jamieh has joined #openstack-marconi10:44
*** flwang has joined #openstack-marconi11:10
*** ykaplan has joined #openstack-marconi11:27
*** ykaplan has quit IRC12:00
*** ykaplan has joined #openstack-marconi12:39
*** haomaiw__ has joined #openstack-marconi12:58
*** vkmc has joined #openstack-marconi12:59
*** haomaiwa_ has quit IRC13:02
*** nosnos has quit IRC13:08
*** malini_afk is now known as malini13:22
*** kgriffs_afk is now known as kgriffs13:26
*** malini is now known as malini_afk13:36
openstackgerritKurt Griffiths proposed a change to openstack/marconi: feat(v1.1): Plumb v1.1 API in the WSGI transport  https://review.openstack.org/7404413:54
*** cpallares has joined #openstack-marconi13:55
flaper87kgriffs: goood morning :)13:55
cpallareso/13:55
*** sriram has joined #openstack-marconi13:57
*** sriram has quit IRC13:57
*** sriram has joined #openstack-marconi13:57
kgriffso/13:59
*** oz_akan_ has joined #openstack-marconi13:59
kgriffsflaper87: I'll be back in an hour or two, but I wanted to get that patch frontloaded so everyone can start reviewing and commenting. ^^^13:59
flaper87kgriffs: hehe, ok14:00
*** mpanetta has joined #openstack-marconi14:05
openstackgerritA change was merged to openstack/marconi: Use six.moves.urllib.parse instead of urlparse  https://review.openstack.org/7282114:10
openstackgerritA change was merged to openstack/marconi: Improve help strings  https://review.openstack.org/7208114:11
*** kgriffs is now known as kgriffs_afk14:15
*** saju_m has quit IRC14:15
*** caleb_ has joined #openstack-marconi14:20
*** jcru has joined #openstack-marconi14:23
*** alcabrera has joined #openstack-marconi14:45
alcabreraGood morning~ :D14:46
*** amitgandhi has joined #openstack-marconi14:46
flwangalcabrera: morning :D14:48
alcabreraflwang: o/14:48
*** amitgandhi has quit IRC14:49
flaper87alcabrera: morniiiiiiiiiiiiiiiiiing14:49
*** amitgandhi has joined #openstack-marconi14:49
flaper87alcabrera: thanks for submitting the talks14:49
flwangalcabrera: did you see the patch from kgriffs_afk https://review.openstack.org/72081 ?14:49
alcabreraflaper87: yep - no problem! :)14:50
alcabreraflwang: not yet. *clicks*14:50
flwangflaper87: alcabrera: do you think we need another v1_1 folder under common.transport.wsgi?14:50
flaper87flwang: we do, kgriffs_afk just submitted a patch for that14:50
flwangflaper87: alcabrera: for the new health?14:50
alcabrerahmm14:51
alcabreraI'll take a look14:51
flaper87flwang: https://review.openstack.org/#/c/74044/ this patch adds the package for v1_114:51
flwangalcabrera: sorry, wrong link  https://review.openstack.org/7404414:51
flwangflaper87: it's just for queues.transport.wsgi14:51
alcabrerawe might need another v1_1 folder, since common.transport.wsgi houses the common Health responder14:51
flwangalcabrera: exactly14:52
alcabreraor14:52
alcabrera...14:52
alcabrerahmmm14:52
alcabrerathe was moved out to common because it was used in the proxy14:52
alcabrerathe proxy is long gone14:52
alcabreraso it might make more sense to move health back to queues14:52
alcabreraand get rid of that folder entirely14:52
alcabrera**that was moved out ...14:53
flwangalcabrera: but do you think health is about queue?14:53
flwangI assume it's about marconi itself14:53
flwangjust think aloud14:53
alcabreraflwang: I agree. I mean - queues used to have a 'proxy' project living inside that used the same health responder as Marconi queues. :)14:54
flwangalcabrera: got it, thanks14:54
flwangalcabrera: so what's your opinion?14:54
alcabreramv marconi/queues/common/transport/wsgi/health.py -> {marconi/queues/transport/wsgi/v1_{0,1}/healthy.py}14:55
alcabrera*health.py14:55
alcabrerathat's what I think the correct solution is here14:55
flaper87check-tempest-dsvm-marconi FAILURE in 8m 33s (non-voting) w00000000000000000000000000000t14:56
flaper87we've a gate :D14:56
alcabrerasweet14:57
flaper87FWIW, I just commented in kgriffs_afk's patch14:57
alcabreramalini_afk: thanks for all your work on tempest!14:57
alcabreraI'll check out the v1_1 patch once I sort through all my morning email. :P14:57
flwangalcabrera: cool, thanks14:58
*** balajiiyer has joined #openstack-marconi14:58
flwangflaper87: did you see the chat between alcabrera and me? do you think it make sense?14:58
flaper87flwang: not sure what the need is15:00
flaper87flwang: 74044 already adds a v1_1 package15:00
flaper87what do you need?15:00
flwangflaper87: let me reword it again15:01
flwangflaper87: we need to provide a new health for v1_1, but the old health.py which located in common.transport.wsgi15:01
alcabreraflaper87: this (https://github.com/openstack/marconi/blob/master/marconi/common/transport/wsgi/health.py) needs to moved and dup'd to queues/transport/wsgi/{version}/15:02
flwangflaper87: so my question is should we create a new folder v1_1 under common.transport.wsgi15:02
openstackgerritYeela Kaplan proposed a change to openstack/marconi: [WIP]Implement Claim controller for sqlalchemy storage backend  https://review.openstack.org/7406815:02
openstackgerritYeela Kaplan proposed a change to openstack/marconi: feat(sqlalchemy): add shards controller  https://review.openstack.org/7133515:02
openstackgerritYeela Kaplan proposed a change to openstack/marconi: [WIP]Implement Queue controller for sqlalchemy storage backend  https://review.openstack.org/7094715:02
openstackgerritYeela Kaplan proposed a change to openstack/marconi: Sqlalchemy Message controller  https://review.openstack.org/7267715:02
openstackgerritYeela Kaplan proposed a change to openstack/marconi: feat(sql/catalogue): add catalogue controller  https://review.openstack.org/7302715:02
openstackgerritYeela Kaplan proposed a change to openstack/marconi: feat(sql/driver): expose ControlDriver, more config  https://review.openstack.org/7020215:02
flwangflaper87: or just move it to queues.transport.wsgi.v1_1 as well?15:02
flaper87mmmh, something seems wrong there15:02
flaper87ykaplan: for some reason gerrit re-submitted alcabrera's patches15:03
flaper87:/15:03
flaper87did you change anything there ?15:03
alcabrerawow, that's some intense rebsing!15:03
alcabrera*rebasing15:03
flwangflaper87: I think ykaplan may miss the -R argument?15:03
ykaplanflaper87, alcabrera: Sorry15:03
ykaplanno15:03
ykaplanI did15:03
flaper87alcabrera: your patches looked quite well last week, I think I commented in one of those, lets get them merged today15:04
ykaplansorry :/15:04
flwangflaper87: am I clear?15:04
flaper87ykaplan: not big deal, he can get the changes and re-submit those patches :)15:05
flaper87alcabrera don't forgive ykaplan, ask for some pop-tarts, gummy bears ANYTHING ;)15:05
* flaper87 hides and runs away15:05
alcabreralol15:05
* flaper87 sees ykaplan is chasing him15:05
alcabreraykaplan: no worries. :)15:05
* flaper87 keeps running for his life15:05
flaper87:D15:05
* alcabrera eats all the poptarts, anyway15:06
*** chandan_kumar has quit IRC15:07
*** rustlebee is now known as russellb15:08
flaper87flwang: ah, mmh! It needs to be copied into v1 and v1_1 then15:09
flwangflaper87: under where?15:09
flwangcommon? or queues?15:09
*** rwsu has joined #openstack-marconi15:09
flaper87queues, I'd say15:09
flaper87alcabrera: ^ ?15:09
alcabreraflaper87, flwang: +115:10
alcabreraqueues15:10
flwangalcabrera: flaper87: cool, get the approval from 2 of 3 cores :D15:10
* flaper87 clones himself 4 times15:11
flaper87flwang: no you've the approval from 2 out of 715:11
flaper87:D15:11
alcabreralol15:12
flwangwho is the him? O:-)15:12
alcabreramap (approve) flapers15:12
*** malini_afk is now known as malini15:16
alcabrerakgriffs_afk: public_endpoints and private_endpoints - nice!15:19
flwangalcabrera: +1, yep, I think this way instead of the admin/public15:20
alcabreramarconi meeting tomorrow - get your items on the agenda: https://wiki.openstack.org/wiki/Meetings/Marconi#Agenda15:23
alcabrera:D15:23
malinio///15:24
maliniflaper87: ping15:24
flaper87malini: pong15:24
maliniflaper87: devstack is bothering:-( me15:25
* flaper87 takes a bat and beats down devstack15:25
flaper87no one messes with malini mmhhh15:25
alcabreralol15:26
alcabreracheck-tempest-dsvm-marconi                      FAILURE in OUCHm BATs (non-voting)15:26
maliniflaper87: sorry had to step away15:27
flaper87alcabrera: LOOOOOOOOOOOOOOOOOOOL15:27
flaper87good one15:27
maliniflaper87: The job is failing coz it cannot connect to mongo15:27
maliniflaper87: http://logs.openstack.org/49/63449/16/experimental/check-tempest-dsvm-marconi/4a2558c/console.html15:28
maliniflaper87: http://paste.openstack.org/show/66512/15:28
maliniany idea what could be going on?15:28
flaper87yes15:29
flaper87that's the first mongodb boot which means it is creating journal files15:29
flaper87you may want to start mongo with --nojournal15:29
flaper87(I think that's the right flag, double check pls)15:29
flaper87so that it doesn't have to pre-allocate journal files and it'll be available right away15:29
flaper87malini: --nojournal                 disable journaling (journaling is on by default ....)15:30
alcabreraah, that makes sense15:30
alcabrerait takes like... 10 seconds to boot mongo when it has to allocate the things15:30
flaper87alcabrera: hehe, 10 seconds if all planets are aligned....15:31
maliniflaper87: shud that be here https://github.com/openstack-dev/devstack/blob/master/lib/marconi#L105 ?15:31
flaper87... or there, yes!15:32
*** jergerber has joined #openstack-marconi15:32
flaper87malini: journal=false15:33
*** ykaplan has quit IRC15:34
*** ykaplan has joined #openstack-marconi15:45
*** whenry has quit IRC15:48
flaper87alcabrera: did you re-submit your patches ?15:51
alcabreranot yet. I'll do that now. :)15:52
alcabreraflaper87: I still need to address your comments in the shards patch15:53
flaper87alcabrera: cool beans!15:56
flaper87we're literally almost there15:56
*** jergerber has quit IRC15:57
alcabrerayes! :)15:57
*** caleb_ has quit IRC16:03
*** jergerber has joined #openstack-marconi16:05
*** ametts has joined #openstack-marconi16:05
*** caleb_ has joined #openstack-marconi16:06
*** ametts has quit IRC16:07
*** whenry has joined #openstack-marconi16:07
*** jamieh has quit IRC16:15
maliniflaper87: the whole nsize statement looks greek to me (& I am not greek)16:15
malinican you update http://paste.openstack.org/show/66534/ to how the journal shud be added16:15
maliniif you want to go ahead & submit the devstack patch , tht is cool too ;)16:16
*** jamieh has joined #openstack-marconi16:17
*** kgriffs_afk is now known as kgriffs16:22
flaper87malini: http://paste.openstack.org/show/66537/16:24
flaper87I tested the fedora one for you16:24
flaper87I *think* ubuntu's case should work16:24
malinithank you +  cherry pop-tarts flaper87!!!16:27
*** oz_akan_ has quit IRC16:27
*** oz_akan_ has joined #openstack-marconi16:28
flaper87malini: my pelasure16:28
*** chandan_kumar has joined #openstack-marconi16:29
*** chandan_kumar has quit IRC16:29
*** chandan_kumar has joined #openstack-marconi16:30
*** caleb_ has quit IRC16:30
*** caleb_ has joined #openstack-marconi16:36
*** ykaplan has quit IRC16:42
*** chandan_kumar has quit IRC16:55
openstackgerritKai Zhang proposed a change to openstack/marconi: Fix two small typos in marconi/queues/storage/sharding.py  https://review.openstack.org/7396917:09
*** amitgandhi has quit IRC17:10
*** jamieh has quit IRC17:10
*** rossk has joined #openstack-marconi17:11
openstackgerritA change was merged to openstack/marconi: Fix two small typos in marconi/queues/storage/sharding.py  https://review.openstack.org/7396917:40
*** amitgandhi has joined #openstack-marconi17:40
*** amit has joined #openstack-marconi17:43
*** amitgandhi has quit IRC17:43
*** amit is now known as Guest5555217:43
*** alcabrera is now known as alcabrera|afk17:47
*** Guest55552 has quit IRC17:47
*** caleb_ has quit IRC18:18
*** amitgandhi has joined #openstack-marconi18:37
*** alcabrera|afk is now known as alcabrera18:39
*** amitgandhi has quit IRC18:41
*** amitgandhi has joined #openstack-marconi18:43
*** raxuanyu has joined #openstack-marconi18:47
*** openstackgerrit has quit IRC18:53
*** openstackgerrit has joined #openstack-marconi18:54
*** amitgandhi has quit IRC18:55
kgriffsalcabrera: ping18:57
openstackgerritKurt Griffiths proposed a change to openstack/marconi: feat(v1.1): Plumb v1.1 API in the WSGI transport  https://review.openstack.org/7404418:58
*** haomaiw__ has quit IRC18:58
*** haomaiwang has joined #openstack-marconi18:59
kgriffsalcabrera: how do you feel about moving the remaining two modules (helpers, utils) from common.transport.wsgi to queues.transport.wsgi ?18:59
alcabrerahmmm19:01
alcabrerasounds good to me, kgriffs19:01
alcabreraotoh, it makes me think of marconi 'notifications'19:02
alcabreraso19:02
alcabrerait depends on whether those functions will help implement the API layer for notifications19:03
kgriffsmmm19:04
kgriffsgood point19:04
kgriffslet's keep them as-is for now19:05
alcabrera+119:05
alcabreraI was about to say. :)19:05
kgriffsI think there's a good chance notifications will wants them19:05
alcabreraload(), validate() - yes, very19:05
*** caleb_ has joined #openstack-marconi19:10
*** vkmc has quit IRC19:18
*** caleb_ has quit IRC19:30
*** amitgandhi has joined #openstack-marconi19:31
*** amitgandhi has quit IRC19:32
*** amitgandhi has joined #openstack-marconi19:32
*** jcru has quit IRC19:49
*** jcru has joined #openstack-marconi19:50
*** kgriffs is now known as kgriffs_afk19:59
*** jay-atl has joined #openstack-marconi20:01
*** kgriffs_afk is now known as kgriffs20:04
*** caleb_ has joined #openstack-marconi20:18
*** malini is now known as malini_afk20:27
*** kaizh has joined #openstack-marconi20:57
*** whenry has quit IRC21:00
kaizhHey, guys! Is is better to print a 'Serving on host %(bind)s:%(port)s' to stdout instand of log after starting the server? so that user can know immediately if server started successfully or blocked by something?21:08
kgriffskaizh: that's a good question21:10
kgriffsare you referring to marconi-server ?21:10
kaizhYes21:10
kgriffskk21:10
kgriffsso, the original implementation actually did that21:11
kgriffsbut then it got changed, and I am trying to remember the reasoning...21:11
kgriffsbtw, you can pass -v I believe, which will cause it to output that string21:11
kgriffs(to stdout)21:11
*** whenry has joined #openstack-marconi21:12
kgriffsflaper87: ^^^21:12
kgriffsflaper87: do you remember the reasoning?21:12
flaper87the original did that because we were using gunicorn21:13
flaper87:D21:13
flaper87but you can get that info by using `-v -d`21:13
flaper87IIRC21:13
flaper87kaizh: ^21:14
kgriffsmy local install is borked at the moment, so I can't verify, but that sounds right. You may not even need the '-d'21:14
kaizhI've tried -v. and yes, it prints the log21:14
flaper87kaizh: you need -d :D21:15
flaper872014-02-17 22:15:04.548 29992 INFO marconi.queues.transport.wsgi.driver [-] Serving on host 127.0.0.1:888821:15
*** sriram has quit IRC21:15
flaper87well, you need INFO level logs21:15
flaper87to be precise21:15
kgriffsflaper87: what do you think about kaizh's suggestion to just print that startup string when not echoing the log to stdout?21:17
kgriffssince, it is really a toy server21:17
kgriffsfor playing around on the cmd line and stuff21:17
kgriffsand the user needs to know the port21:17
kgriffsso, the idea is to print out "serving on ...." when '-v' is not passed21:18
kgriffsor something like that.21:18
flaper87mmh, TBH, I'd prefer to stick to -v. Not fully opposed to the proposal but the reasoning is that it'll keep everything consistent with other parts of marconi.21:22
flaper87I mean, it kinda feels wrong to not allowing people to shut marconi's output off, even in the toy server21:23
flaper87but I could eb wrong so, not fully oppossed to the proposal21:23
alcabreraI remember something about marconi-server encountering some errors in the context of stdout not being available21:23
kgriffsbtw guys, i updated that patch and responded to flaper87's comments21:26
* flaper87 opens the review just because he's curious to read kgriffs answers21:26
flaper87... and because I'm addicted to this stuff21:26
flaper87damn21:27
kgriffscuriosity killed the coder21:27
flaper87kgriffs: >= (current - 2) <- we're on the same page here21:28
flaper87I fully agree with your reasoning there21:28
flaper87kgriffs: "Are other projects advocating now for this new direction?" <- I have to say it: muahahahhaha muahahahah muahahahhahahahaha21:29
flaper87:D21:29
flaper87Shall we be the first?21:29
kgriffsheh21:29
flaper87jokes apart21:30
kgriffsso, I heard something about keystone doing something with versioning21:30
kgriffsbut21:30
kgriffsI know next to nothing about it21:30
flaper87I haven't heard of any yet but I'd really like us to consider it for June21:30
flaper87at least have a deep discussion about that21:30
flaper87with pros / cons etc21:30
kgriffsI think it would be worth a summit discussion with a couple other project teams21:31
flaper87perhaps a design session ?21:31
flaper87We can invite keystone guys21:31
kgriffsflaper87: yeah, if they do create a cross-project track21:31
flaper87indeed21:31
kgriffsthen that would be perfect21:31
flaper87hopefully there will be one21:31
kgriffsso, there are two things we need to sort out for marconi:21:32
alcabreraI'm out for the night. Take care, guys. :)21:32
kgriffs1. how to do versioning post-1.0 (this is the session we are talking about)21:32
kgriffs2. whether operators should be able to turn off versions or it is hard-coded that you must support at least x back from current21:33
kgriffsunless there is a really compelling reason to let operators do that, for the sake of the user, I would prefer having a hard-coded set21:33
*** alcabrera has quit IRC21:34
kgriffswrt #2 are there other projects you are aware of that allow that?21:34
kgriffs(via stevedore or something)21:34
kgriffsflaper87: btw, I think I remember hacking the logging level to be info just for that initial string, then it reverts to whatever it was before the hack21:36
kgriffshmm... maybe not21:37
* kgriffs is looking through the code21:37
kgriffsoh, nevermind21:38
kgriffsthat's what -v does21:38
kgriffs:p21:38
flaper87:)21:38
*** cpallares has quit IRC21:39
*** cpallares has joined #openstack-marconi21:41
openstackgerritKurt Griffiths proposed a change to openstack/marconi: doc(README): Have the user pass '-v' when starting marconi-server  https://review.openstack.org/7417221:42
*** kaizh has left #openstack-marconi22:08
*** oz_akan_ has quit IRC22:20
*** amit has joined #openstack-marconi22:22
*** amit has quit IRC22:22
*** amit has joined #openstack-marconi22:23
*** amit is now known as Guest351522:23
*** amitgandhi has quit IRC22:24
*** kgriffs is now known as kgriffs_afk22:25
*** balajiiyer1 has joined #openstack-marconi22:26
*** balajiiyer1 has quit IRC22:27
*** balajiiyer has quit IRC22:30
*** cpallares has quit IRC22:50
*** mpanetta has quit IRC22:55
*** kgriffs_afk is now known as kgriffs23:08
*** caleb_ has quit IRC23:10
openstackgerritA change was merged to openstack/marconi: feat(sql/driver): expose ControlDriver, more config  https://review.openstack.org/7020223:16
*** sriram has joined #openstack-marconi23:17
openstackgerritA change was merged to openstack/marconi: Apply six for metaclass  https://review.openstack.org/7313523:20
*** jergerber has quit IRC23:37
*** Guest3515 has quit IRC23:40
*** mwagner_lap has quit IRC23:51
*** jcru has quit IRC23:52
*** jcru has joined #openstack-marconi23:53
*** sriram1 has joined #openstack-marconi23:56
*** sriram has quit IRC23:57
*** jcru has quit IRC23:57
« Sunday, 2014-02-16 Index Tuesday, 2014-02-18 »

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