Monday, 2014-03-03

« Sunday, 2014-03-02 Index Tuesday, 2014-03-04 »
*** nosnos has joined #openstack-marconi01:28
*** haomaiwang has quit IRC02:22
*** haomaiwa_ has joined #openstack-marconi02:23
*** cpallares has joined #openstack-marconi02:55
*** cpallares has quit IRC02:55
*** flwang has quit IRC03:10
*** flwang has joined #openstack-marconi03:47
*** nosnos has quit IRC04:28
*** nosnos has joined #openstack-marconi04:34
*** haomaiwa_ has quit IRC05:48
*** haomaiwa_ has joined #openstack-marconi05:49
*** nosnos has quit IRC06:01
*** nosnos has joined #openstack-marconi06:06
*** saju_m has joined #openstack-marconi06:31
*** nosnos has quit IRC06:40
*** chandankumar_ has quit IRC06:45
*** chandan_kumar has joined #openstack-marconi06:51
*** chandan_kumar has quit IRC06:55
*** nosnos has joined #openstack-marconi07:04
*** haomaiwang has joined #openstack-marconi07:12
*** haomaiwa_ has quit IRC07:15
*** davidhadas has quit IRC07:35
*** flaper87|afk is now known as flaper8708:00
flwangflaper87: morning08:16
flaper87flwang: hey hye08:17
flwangflaper87: do we have to land all bp patches asap given the FF is coming?08:18
flaper87well, just if they're related to a blueprint08:18
flaper87I mean, if they introduce a new feature08:19
flaper87otherwise no08:19
flwangI understand, so I'm afraid of my /health bp08:24
flaper87flwang: don't be, it's healthy :P08:25
flwangflaper87: I have not added test cases given the sqlalchemy change and the v1.1 test change08:25
flaper87flwang: we need to discuss a bit how we wan't to land all those patches. I mean, I don't think v1.1 should / will be part of Icehouse so we propably want to defer some patches08:25
flwangflaper87: sounds good08:26
flwangso please take a look about the overall shape when you're available, I have addressed you and kgriffs_afk's comments08:27
*** davidhadas has joined #openstack-marconi08:27
flwangI think we can discuss that on tomorrow meeting08:28
flwanghttps://review.openstack.org/#/c/76555/08:29
*** haomaiwang has quit IRC08:35
*** haomaiwang has joined #openstack-marconi08:35
*** davidhadas_ has joined #openstack-marconi08:36
*** davidhadas has quit IRC08:39
*** saju_m has quit IRC08:40
*** saju_m has joined #openstack-marconi08:42
*** saju_m has quit IRC09:02
*** chandan_kumar has joined #openstack-marconi09:03
*** flwang has quit IRC09:13
*** saju_m has joined #openstack-marconi09:34
*** jamieh has joined #openstack-marconi09:50
*** jamieh_ has joined #openstack-marconi10:03
*** flwang has joined #openstack-marconi10:04
*** jamieh has quit IRC10:05
*** ykaplan has joined #openstack-marconi10:09
*** jamieh_ has quit IRC10:19
*** jamieh has joined #openstack-marconi10:22
*** saju_m has quit IRC10:25
*** jamieh has quit IRC10:25
*** jamieh has joined #openstack-marconi10:31
*** saju_m has joined #openstack-marconi10:41
*** ykaplan has quit IRC11:03
*** saju_m has quit IRC11:11
*** nosnos has quit IRC11:21
*** saju_m has joined #openstack-marconi11:25
*** ykaplan has joined #openstack-marconi11:26
*** flwang has quit IRC11:28
*** malini_afk is now known as malini11:30
*** flwang has joined #openstack-marconi11:44
*** ykaplan has quit IRC11:58
*** jamieh has quit IRC12:07
*** jamieh has joined #openstack-marconi12:08
*** ykaplan has joined #openstack-marconi12:16
*** seiflotfy has joined #openstack-marconi12:29
seiflotfyhey guys12:29
seiflotfyflaper87 hey12:31
seiflotfyaround?12:31
seiflotfyi wanted to know if its ok for me to pathc up https://bugs.launchpad.net/marconi/+bug/122293212:31
seiflotfy?12:31
*** malini is now known as malini_afk12:37
*** chandan_kumar has quit IRC12:48
*** chandan_kumar has joined #openstack-marconi12:49
*** saju_m has quit IRC12:52
*** chandankumar_ has joined #openstack-marconi12:54
flaper87seiflotfy: hey hey12:54
flaper87I am now12:54
flaper87seiflotfy: yeah, please!12:55
flaper87seiflotfy: ah wait12:55
flaper87seiflotfy: flwang is assigned to that12:55
flaper87flwang: are you working on that bug?12:55
*** chandan_kumar has quit IRC12:57
seiflotfyflaper87i intend to use marconi for my own stuff but as a standalone12:58
seiflotfyits good stuff12:58
seiflotfyflaper8713:00
flaper87seiflotfy: I'm happy to hear that! Please, feel free to share any thoughts, feedback and / or complains here or on the mailing list!13:02
flaper87we're looking forward to improve it as much as possible13:02
*** nidhi has joined #openstack-marconi13:05
flaper87seiflotfy: may I ask what your usecase is?13:07
seiflotfywell i use rabbitmq for github.com/seiflotfy/compatipede13:08
seiflotfywhich is used by mozilla internally13:09
seiflotfyi intend to move it away from rabbitmq and marconi13:09
seiflotfythere is no reason i am doing that really13:09
seiflotfyjust feels fun to help out on something awesome13:09
seiflotfyi do prefer my technology to all be in python13:09
flwangflaper87: yes, after a long ^run^ with Glance, I'm going to pay more attention on Marconi, so the bug 1222932 is on my list13:21
flwangseiflotfy: if you're finding some low-hanging-fruit, I think there are a lot in Marconi13:22
flwangseiflotfy: for example, you can add test cases for Marconi13:22
seiflotfyyeah13:22
seiflotfybut not so motivating to right test cases13:22
*** jamieh has quit IRC13:23
flaper87seiflotfy: awesome, pls, let us know what you think13:25
flaper87and what your experience is13:25
flaper87I'm quite interested to hear what the integration process is like13:25
flaper87as in, downloading marconi and using the client13:25
flaper87seiflotfy: what did you use to talk to rabbitmq? kombu?13:26
*** jamieh has joined #openstack-marconi13:30
seiflotfyflaper87 pika13:33
seiflotfyflaper87 do u have any bugs you would like me to look at besides test cases (not motivated to do that)13:34
*** tedross has joined #openstack-marconi13:35
*** westmaas has quit IRC13:38
*** westmaas has joined #openstack-marconi13:39
*** dragondm has quit IRC13:40
*** dragondm has joined #openstack-marconi13:40
*** sballe_ has joined #openstack-marconi13:44
*** sballe has quit IRC13:48
nidhiHi I built marconi from the github source code13:50
nidhihttps://github.com/openstack/marconi13:50
nidhithe server seems to be running fine13:50
nidhibut the queue creation gives a HTTP 503 error13:51
*** rustlebee is now known as russellb13:56
flwangnidhi: what's the trace?13:56
*** nidhi has quit IRC14:00
*** cpallares has joined #openstack-marconi14:17
cpallaresgood morning!14:18
flwangcpallares: morning :)14:19
cpallareshello flwang, how are you doing?14:19
flwangcpallares: good, just reviewed your patch and left my 2 cents14:19
cpallaresflwang, thanks. I appreciate that.14:20
flwangcpallares: my pleasure14:20
*** mpanetta has joined #openstack-marconi14:32
*** ykaplan has quit IRC14:39
*** nidhi has joined #openstack-marconi14:49
*** jergerber has joined #openstack-marconi14:49
flaper87ops, looks like my irc notifications failed me14:50
flaper87seiflotfy: I'll take a look, there are some you may be interested in14:50
flaper87nidhi: any traceback? Could you try running $ marconi-server -d -v ? What are you using to talk to marconi (client lib or curl) ?14:51
nidhii just logged in now14:52
nidhii dont have the previous messages14:52
nidhii am using curl14:52
flaper87nidhi: what about giving the client library a try ? :)14:53
nidhii'll try that14:53
flaper87nidhi: before you yell at me, I know there's no much documentation :(14:53
flaper87nidhi: https://github.com/openstack/python-marconiclient/14:53
flaper87nidhi: but there are examples: https://github.com/openstack/python-marconiclient/blob/master/examples/simple.py14:54
*** amitgandhi has joined #openstack-marconi14:54
*** amitgandhi has quit IRC14:54
*** amitgandhi has joined #openstack-marconi14:55
*** kgriffs_afk is now known as kgriffs14:57
*** Kuwagata has joined #openstack-marconi14:59
nidhiflapper87: what is the command to run marconi server with some other port number ?14:59
flaper87nidhi: as of now, you need to change it in the config file since there's a bug w/ how config options are registered that I'm fixing now15:04
*** jay-atl1 has quit IRC15:08
kgriffsflaper87: did that kombu patch land?15:09
*** malini_afk is now known as malini15:16
flaper87kgriffs: not yet15:23
flaper87kgriffs: I'm a bit afraid to let it land as is since there's a chance for race conditions there15:23
flaper87since the client lib still doesn't have support for claims15:23
flaper87I'd like to first let the claim patch land15:23
flaper87then do a 0.1 release15:24
malinigood morning!!15:25
flaper87malini: good morning!15:25
flaper87malini: does the gate work?15:25
flaper87:D15:25
malinino :(15:26
maliniBut I think we are really close..15:26
maliniclose by an inch, but the inch keeps moving every time :(15:26
maliniThere is something funny going on with the logging15:26
flaper87shoot15:27
flaper87lets get that fixed15:27
maliniif I direct the output to something silly like > /dev/null in https://github.com/openstack-dev/devstack/blob/master/lib/marconi#L151 , it works15:27
maliniI am wondering if we are still logging to console or if its a permissions thing15:28
*** alcabrera has joined #openstack-marconi15:28
alcabreraGood morning! :D15:29
flaper87alcabrera: GOOOOOOOOOOOOOOOOOOOOOD MORNINNNNNNNNNNNNNNNGGGGGGGGGGGGGGGG15:31
flaper87malini: it may be a permissions thing15:31
flaper87malini: the logging lib we're using is the same used by other openstack projects15:32
flaper87meaning, if it doesn't fail for them, it shouldn't fail for us15:32
alcabreraflaper87: yay! How are you? :)15:32
maliniflaper87: I compared with other devstack projects & none of them have anything special for logging in their scripts15:32
flaper87I do think we should fix the logs folder name since our default folder is `marconi` not `marconi-api`15:32
flaper87malini: do we have something special?15:33
maliniflaper87: I did try tht in my local devstack15:33
malinitht didnt help15:33
flaper87alcabrera: all good my friend, you?15:33
maliniflaper87: But I think that is ALSO needed (marconi instead of marcobi-api)15:33
alcabrerafantastic, and only slightly sleepy.15:33
* kgriffs waives15:37
kgriffssoooooooo15:37
alcabrerakgriffs: o/15:37
kgriffswe have less than 48 hours to finish up i-315:37
alcabreramuch review, such time-constraint15:37
kgriffsI am planning on moving out all "low" priority work items that aren't already completed, unless anyone objects15:38
maliniflaper87: I am going to dig more into the permissions thing now15:38
alcabreraI'll be back in a bit. Meeting. :P15:38
*** alcabrera is now known as alcabrera|afk15:38
kgriffsalcabrera|afk: ok15:38
flaper87so, we need to close blueprints but we can still work on bugs :)15:38
kgriffsflaper87: ah, OK. we can allow critical bug fixes in until Wed, right?15:39
flaper87the freeze is for new features not bugs15:41
*** chandan_kumar has joined #openstack-marconi15:41
flaper87so, as long as patches don't add new features, it's fine15:41
kgriffshence the name "FeatureFreeze"15:42
kgriffs:D15:42
flaper87:)15:42
kgriffsdo they cut an i-3 on thursday?15:42
kgriffshmmm15:42
kgriffshttps://wiki.openstack.org/wiki/Icehouse_Release_Schedule15:42
kgriffsLooks like icehouse RCs start 27th15:43
flaper87kgriffs: yup15:43
kgriffsso I am assuming i-3 goes out on the 6th15:43
kgriffsthen we have a couple weeks to keep fixing teh bugz for Icehouse release15:43
flaper87correct15:44
flaper87*just bugs*15:44
flaper87no features15:44
flaper87my sqlalchemy patches would be fine because those fix bugs15:44
flaper87also, we need to ask a graduation review15:44
kgriffsflaper87: yep15:44
flaper87kgriffs: Next week ?15:45
kgriffsflaper87: I thought grad reviews happened just before PTL elections?15:45
kgriffsflaper87: we can ask for next week, but they may ask us to wait another few weeks15:46
kgriffsflaper87: I'd like to make sure our docs are good to go first, in any case15:46
kgriffsalso we need the final report on the Pecan evaluation15:47
flaper87kgriffs: it's fine if they ask us to wait, I just want to make sure we / they don't forget15:47
flaper87:D15:47
kgriffsyep15:47
kgriffslet's talk about graduation tomorrow at our team meeting15:48
kgriffswe can send a note to the TC if everyone agrees we will be ready or almost ready by the end of the week15:49
flaper87kgriffs: kk15:50
kgriffsflaper87: I am going to try to get v1.1 done but if we can't get things reviewed and merged fast enough, I think it would be OK to talk about "experimental v1.1 API support" being available, in our messaging at the Atlanta Summit.15:52
flaper87kgriffs: could you please go through the sql patches again?15:53
kgriffsflaper87: yes15:53
flaper87LOL15:53
flaper87dude, dunno whether we should reading our minds or doing it more often15:53
kgriffslet's all try to get fast turnaround on patches15:53
* flaper87 reviews15:53
kgriffsflaper87: d00d, I'm starting to wonder if *I'm* human now. What have you done to me!15:55
* kgriffs thinks this could be an excellent plot for a movie.15:55
kgriffsflaper87: btw15:56
kgriffsI was going to make retries_on_autoreconnect configurable15:56
kgriffsBUT15:56
kgriffswe don't want to use global configs15:56
cpallareslol15:57
kgriffswait...15:57
cpallareshello flaper87, kgriffs!15:57
kgriffsI think I have an idea15:57
kgriffslet me try this and resubmit that patch, see what you think15:57
kgriffscpallares: o/15:57
kgriffscpallares: saw your Falcon patch - looks good!15:57
cpallareskgriffs: :D15:58
kgriffsjust waiting for alcabrera|afk to review it15:58
kgriffscpallares: are you back in Dallas?15:58
cpallareskgriffs: yep, it snowed here15:58
kgriffssrsly?!15:58
kgriffsAnd I thought it was cold here!15:58
cpallareskgriffs: yeah, there was just ice falling at some point15:59
flaper87cpallares: helloooooooooooooooooooooooooo15:59
cpallareskgriffs: how was the conference?15:59
kgriffsflaper87: how's the weather over there?15:59
kgriffscpallares: it was TEH AWESOME15:59
cpallareskgriffs: haha15:59
kgriffsI heard from lots of people who made new connections and got tons of good ideas that they can't wait to start working on16:00
flaper87kgriffs: this was quite a weird winter... It didn't snow that much and spring is basically already here16:00
flaper87lot of rain16:00
kgriffsah16:00
cpallareskgriffs: Nice. I saw that you got an award? Congratulations :D16:00
kgriffscpallares: thanks16:00
*** jmckind has joined #openstack-marconi16:00
kgriffswas sort of embarrassing, but also sort of awesome16:00
cpallaresflaper87: I'm jealous. Weather here doesn't make up its mind. It's nice, it's cold, it snows, it's hot.16:01
kgriffslol16:01
kgriffswould be nice if we had some random cool days in, say, July16:01
cpallareskgriffs: that would be glorious16:01
kgriffsbut not *this* cool16:01
cpallareshaha16:02
kgriffsflaper87: based on your earlier comments, I will move this out to Juno16:02
kgriffshttps://blueprints.launchpad.net/marconi/+spec/marconi-kombu-support16:02
flaper87kgriffs: wait16:03
flaper87gimme 2 secs, call16:03
*** chandan_kumar has quit IRC16:03
kgriffscpallares: it was great to hang out with you in Austin. Thanks for coming around!16:04
kgriffscpallares: did you ever find your phone?16:04
cpallareskgriffs: Thanks for showing me around!16:04
cpallareskgriffs: Yeah, Anne mailed it back to me, it was in her car.16:04
kgriffsw00t16:04
kgriffscpallares: next time we should do a hardware hackday or something. You can teach me the ways of the EE ninja.16:05
cpallareskgriffs: Kinda disappointed since I wanted another one :P16:05
kgriffsLOOOOL16:05
kgriffsI know how that goes!16:05
cpallareskgriffs: That sounds fun!16:05
cpallareskgriffs: We should16:05
kgriffsI just bought my first soldering iron a few weeks ago16:05
cpallareskgriffs: I might head down there for the next hackathon16:05
* kgriffs is so proud16:05
cpallareskgriffs: Oh neat :D16:05
cpallareskgriffs: You should check out the tech shop!16:06
cpallareskgriffs: They have really neat stuff there.16:06
kgriffswhat's/where's that?16:06
cpallareskgriffs: It's really north, like almost hitting Roundrock16:06
cpallareskgriffs: or it might be in round rock.16:06
malinikgriffs: soldering iron was the reason I wanted to do electrical engineering :D16:07
cpallaresmalini: Are you an EE? :D16:07
maliniyes16:07
cpallaresmalini: wooh16:07
kgriffscpallares: is this what you are talking about? http://techshop.ws/austin_round_rock.html16:07
* cpallares high-fives malini16:07
cpallareskgriffs: yeah16:07
kgriffsmalini: I forgot abou that!16:07
malinibut my sister made more holes in my mom's curtains than me :D16:07
kgriffswe TOTALLY need to get everyone together for a hardware hackday16:07
malinicpallares: are you an EE too ?16:07
cpallaresmalini: yep :D16:08
cpallareskgriffs: I have a coupon for a free one month membership that I never used, next time I see you I'll give it to you.16:08
malinicpallares: yayy!16:08
kgriffscpallares: you are too kind!16:08
cpallareskgriffs: they have 3D printers and oscilloscopes, really neat soldering irons, frequency counter, all the good stuff.16:10
cpallareskgriffs: I think they also give classes.16:10
*** reed has quit IRC16:11
*** chandan_kumar has joined #openstack-marconi16:16
mpanettaHardware!!!16:21
mpanettaI, in my stargate obsession, got an ancient font up on the free sample VFD I got from noritake.16:22
mpanettaI can't believe I missed at talk about hardware... My fav thing ever.16:26
cpallaresmpanetta: are you an EE too? :)16:27
mpanettaI was born one :P  So to say.  Never had any real formal training, but my dad was one back in the day and I have always been around soldering irons or wire wrap tools. :P16:29
mpanettaI don't wirewrap too much anymore, since places like OSHPark make having PCB's done so easy and cheap.16:30
maliniflaper87: can you check if our logging writes to both the log_file and the console (or somewhere else) ?16:31
mpanettacpallares: https://plus.google.com/108215088092353345562/posts/hCxBWRTQqgA  I did that over the weekend.16:32
* mpanetta loves getting free parts to play with16:33
cpallaresmpanetta: that is cool :D is it using the stargate font?16:33
kgriffsI think we should do a hardware hacking meetup in Atlanta around the summit16:33
mpanettacpallares: I hand coded it, but yeah.16:33
cpallareskgriffs: +116:33
mpanettaIt is running off an arduino16:34
mpanettakgriffs: Yes!16:34
kgriffsInternet of Things Powered by OpenStack16:34
mpanettaNow that I can go :P16:34
mpanettaWell I hope.  I guess I should ask my boss first, :P16:34
kgriffsheh16:35
mpanettaI should make a display that I can push messages to using queues.16:35
kgriffsmpanetta: I'm thinking we should all meet on the Saturday before for a hackday or something. Maybe we can even get someone to sponsor it and make it official16:36
kgriffsor...16:36
kgriffs[your idea here]16:36
mpanettakgriffs: Well we have space for that now16:37
cpallareskgriffs: There's lot of neat hardware related stuff during sxsw, you should check it out. Last year, I took my little brothers. There was a makey makey banana piano, and they were giving out little lilypads to make patches.16:37
kgriffsoh, fun16:37
cpallareskgriffs: They give out lots of stuff.16:37
mpanettaI have always wanted to go to SXSW16:37
malinicpallares: we are planning on a kids area in the office..tht sounds like a cool idea for it :)16:37
KuwagataThere are some hackerspaces around Atlanta right?16:37
mpanettaKuwagata: Yes16:38
mpanetta2 at least16:38
mpanettaMore like 3 I think16:38
KuwagataCould probably try to organize something with them maybe.16:38
mpanettaThe Gwinnett Geeks finally have a place now...16:38
cpallaresmalini: you should :) my little brothers (both 11 years old) really enjoyed it, but I saw little kids there.16:38
mpanettaSparkfun sells fiberoptic cloth now16:39
*** balajiiyer has joined #openstack-marconi16:39
mpanettaKuwagata: We could.16:40
*** alcabrera|afk is now known as alcabrera16:40
cpallaresmpanetta: is this it? https://www.sparkfun.com/products/12712 It looks aweeeesome.16:40
alcabrerahey, guys. Let's see what I've missed. :P16:40
mpanettaWorst comes to worst I have all the equipment that anyone would ever need.  Including some oscilloscopes lol16:40
* alcabrera catches up16:40
mpanettacpallares: Yep that is the one :)16:40
mpanettaI bet the wandboard quad could run some of openstack.  I wish it had one of the ARM's on it that supported virtualization though :(16:42
mpanettaI want a NON-X86(_64) cluster.  I think that would be way cool.16:42
kgriffsalcabrera: we are discussing partnering with Gwinnett Geeks or something to sponsor an OpenStack-themed Internet of Things hackday the saturday before the summit or something16:45
* flaper87 back16:47
flaper87kgriffs: re kombu16:47
flaper87I don't think we should move it to icehouse nor that it is tight to the FF16:47
flaper87it won't land in our codebase16:47
kgriffsoic16:47
kgriffswell...16:47
kgriffsI know ttx will ask about it16:47
flaper87actually, it doesn't make much sense to have that bp other than just reminding us that we have to do it16:48
kgriffsright16:48
*** davidhadas_ has quit IRC16:52
*** ykaplan has joined #openstack-marconi16:55
*** jay-atl has joined #openstack-marconi16:55
kgriffsflaper87: just got a good use case for pop functionality17:05
kgriffssomeone wants to consume loglines, perf data and they are ok if they lose a few messages once in a blue moon (when a worker dies)17:05
kgriffsplus, they want multi-pop17:06
kgriffslike, pop 10-50 messages at a time17:06
kgriffsflaper87: did that ever get added to the api 1.1 blueprint?17:06
kgriffsalcabrera: ^^^17:06
kgriffsflwang: ^^^17:06
* alcabrera catches up again17:10
alcabreraI got pulled into an awesome discussion locally17:11
alcabreralet's see, for real now17:11
alcabreraI caught on that malini has a background in EE (awewsome)17:11
alcabreraglad to hear that you found your phone, cpallares. :)17:11
alcabrerampanetta does all the hardware17:11
alcabreraand we need a graduation review in the next week17:11
* alcabrera feels caught up17:11
mpanettaNot all :P17:12
alcabreralol17:12
cpallaresalcabrera: haha thanks17:18
cpallaresalcabrera: how was your trip/presentation/the conference/your weekend?17:18
alcabrerait was *amazing*!17:19
alcabreraI was really happy with how the workshops I was teaching turned out, in particular17:21
alcabreraI didn't end up doing all the teaching. :P17:21
alcabreraThe attendees kicked in and we worked together to explain Monads and currying over in Haskell land, and then during the "Python, Openstack, and You" session -17:21
alcabreraMany, many great questions and stories were shared about working on/with openstack.17:21
alcabreraI couldn't have asked for a better turn out!17:21
*** nidhi has quit IRC17:22
cpallaresalcabrera: that sounds pretty awesome :)17:23
*** chandan_kumar has quit IRC17:25
*** cpallares has quit IRC17:28
kgriffsalcabrera: I need to chat with you some time about that second session17:29
kgriffsI'd like to cross-reference with the community unconference session I lead17:29
*** davidhadas has joined #openstack-marconi17:29
alcabrerakgriffs: sounds great!17:30
*** cpallares has joined #openstack-marconi17:30
alcabreracpallares: yeah. :D17:30
kgriffsmalini: what's the latest news on tempest?17:33
kgriffsany progress on the logging issue?17:33
malinikgriffs: there is progress, but we havent progressed enough :|17:34
maliniit looks like there is something going on with our logging17:34
alcabrerait's a battke against stdout, kgriffs - wrt to devstack patch17:34
alcabrera*battle17:34
maliniI have a suspicion we are logging to stdout for some reason17:34
*** amitgandhi has quit IRC17:35
malinikgriffs, alcabrera: this is my current suspect https://github.com/openstack/marconi/blob/master/marconi/openstack/common/log.py#L52317:36
*** amitgandhi has joined #openstack-marconi17:37
maliniwe might be missing a config ?17:37
malinistill checking17:37
alcabreramalini will discover the root cause!17:39
alcabreraI mean, seriously17:39
alcabrerathanks so much for digging on this all this time17:39
alcabreraI'd have lost all my hair working through this. :P17:39
alcabreraIt sounds infinitely frustrating.17:39
maliniI am losing sleep over this17:39
maliniseriously17:39
alcabrera:(17:39
maliniI am waking up at 2 & 3 wondering what this is :-(17:39
*** ykaplan has quit IRC17:40
maliniSo this weekend I decided to NOT look at it & it was tough17:40
alcabreradang, malini17:41
alcabreraso I'm calling this an official openstack pain point17:41
alcabreraputting it in my little black book. :P17:41
*** amitgandhi has quit IRC17:42
maliniI think the problem is we keep looking at how other projects implemented stuff. We really need templates we can refer to, instead of other projects :(17:44
kgriffsalcabrera: you should rename your PDF: "Contributing to OpenStack: The Little Black Book" ;)17:44
kgriffsmalini: so, without a config, logging may default to stdout, is that what you suspect?17:45
alcabrerakgriffs: hahaha17:45
kgriffsalcabrera: speaking of which, I want to contribute to that project17:45
malinikgriffs: or we might be missing a config in logging.conf17:46
kgriffsI'd like to have a chapter about community dynamics/politics.17:46
kgriffsmaybe there is one already? I haven't looked too closely yet at what you have so far17:46
kgriffsmalini: ok, so there is a logging.conf already in the devstack thing?17:46
maliniI though alcabrera was writing the 'Little' black book ;)17:47
alcabrerakgriffs: there isn't one yet, and it'd be a super welcome asddition17:47
alcabreramalini: yeah, that'd be the side project! :D17:47
alcabreraIt'd be a blog where I grump for days17:47
alcabrerathen, at the end of the week, I'd say, "Sure, I'm grumpy. Here's what we can do!"17:48
malinikgriffs: no, but other projects dont write out a logging.conf either..I am trying to figure out waht is different b/w us & other projects17:48
alcabrerado we run wsgiref and other projects use something else?17:48
malinikgriffs, alcabrera, flaper87: Any idea which file this gets populated from https://github.com/openstack/marconi/blob/master/marconi/openstack/common/log.py#L190 ?17:51
kgriffsmalini: you mean, which *.conf file?17:52
maliniyes17:52
alcabreramalini: it should be populated from the application that uses this. So, Marconi should have a "log_file" entry in marconi.conf17:52
alcabrerahttps://github.com/openstack/marconi/blob/master/marconi/openstack/common/log.py#L23017:53
alcabrerabecause we do cfg.CONF(project='marconi', ...)17:53
maliniI am running pdb & the log_file is showing up as 'None' though I have log_file defined in marcobi.conf17:53
alcabrerain marconi.queues.transport.wsgi.app17:53
alcabrerahmmm...17:53
maliniCan somebody else give it a try?17:53
* alcabrera can't be that some body at the moment - hands tied. D:17:54
kgriffsthat could be the problem17:55
kgriffshttps://gist.github.com/anonymous/2e6759b3a36bb0cdca4b17:55
*** rwsu has joined #openstack-marconi17:55
malinikgriffs: yeah17:55
kgriffssys.stdout becomes the default17:55
malinican u try it? just want to make sure its not my conf file17:55
kgriffsI used to think there was a default log path - maybe that changed at some point with oslo logging?17:56
kgriffsmalini: ok17:56
openstackgerritKurt Griffiths proposed a change to openstack/marconi: fix(MongoDB): Driver does not retry on AutoReconnect errors  https://review.openstack.org/7516517:58
kgriffsmalini: I created ~/.marconi/marconi.conf18:01
kgriffsset logfile18:01
kgriffsand _get_log_file_path read it correctly18:01
malinikgriffs: did you set logfile or log_file ?18:01
kgriffslog_file18:01
maliniunder DEFAULT ?18:01
kgriffsyes18:02
kgriffshttps://gist.github.com/anonymous/ba291a38d6177f2387bf18:02
kgriffsI will try use_syslog18:02
malinihmm...18:03
*** ykaplan has joined #openstack-marconi18:04
openstackgerritKurt Griffiths proposed a change to openstack/marconi: fix(MongoDB): Driver does not retry on AutoReconnect errors  https://review.openstack.org/7516518:05
*** amitgandhi has joined #openstack-marconi18:06
*** ykaplan has quit IRC18:07
*** amitgandhi has quit IRC18:07
*** amitgandhi has joined #openstack-marconi18:07
kgriffsI wonder if we need to set use_stderr = False18:08
kgriffsor is it just stdout that is causing the problem?18:08
mpanettaI didn't even know that was an available option...18:09
mpanettaIt may be both18:09
kgriffsok18:09
mpanettaThere is all sorts of weird redirecting of FD's in devstack...18:09
kgriffsuse_syslog defaults to False, so we shouldn't need to set that explicitly18:09
mpanettaIt is like trying to chase Allice's rabbit down the rabbit hole heh18:09
kgriffsmpanetta: EPIC FAIL18:09
mpanettakgriffs: hmm?18:10
kgriffsall kinds of undocumented FD redirecting18:10
mpanettaNot my fault!18:10
mpanetta:P18:10
mpanettaYeah18:10
mpanettaHave you seen it?18:10
kgriffsnope, just going off what you said18:10
mpanettaAh, you should take a look some time.  heh18:10
kgriffsmalini: so, maybe we need these in the marconi.conf ?18:11
mpanettaJamie, Malini and I were looking at it last friday.18:11
kgriffslog_file = /var/log/marconi/queues.log18:11
kgriffsuse_stderr = False18:11
malinikgriff:trying tht18:11
kgriffsof course, that marconi dir will need to be created18:11
mpanettakgriffs: I don;t think /var/log is good18:11
mpanettaIt has to be writeable by stack user18:11
kgriffsmpanetta: wherever we wantz it18:11
mpanettaAh yeah18:12
*** amitgandhi has quit IRC18:12
mpanettaAm I allowed to say that devstack is a mess?  Or will I get kicked in the junk for that? :P18:12
*** amitgandhi has joined #openstack-marconi18:12
malinikgriffs, mpanetta: when I do 'marconi-server &' I get some stuff printed out to my console18:13
maliniDoes tht sound ok?18:13
mpanettaYeah, that is the logging stuff I think18:13
mpanettaIt is OK with me, but maybe not with devstack?18:13
kgriffsmpanetta: you are allowed to say it18:13
kgriffsI don't think it is news to anybody. :p18:13
mpanettakgriffs: Ah I was not sure. ;)18:14
maliniI feel we are really really close18:14
mpanettamalini: I think the issue is it prints stuff every time you connect and do a queue operation...18:14
kgriffsjraim & Co. did a hackday project last friday to start fixing that18:14
malinithis is frustrating18:14
mpanettakgriffs: Oh sweet18:14
mpanettaThe problem is, I am not sure it is ok to print things when screen is not running.18:15
malinimpanetta: 'I am not sure it is ok to print things when screen is not running.' I think that is a problem18:15
kgriffsmalini: why would it print if logging is set? Won't that prevent the stdout logging handler from being configured?18:15
maliniwhere wud it print to?18:15
mpanettamalini: To the log you told it to, I hope. ;)18:16
mpanettaApparently it is doing that, and more...18:16
*** rossk has joined #openstack-marconi18:16
kgriffsAs long as we seet log_path in the config, it shouldn't install the sys.stdout logger18:16
kgriffsI did a quick test, ran marconi-server -v, and it didn't log to console18:17
malinilog_path ?18:17
mpanettawhats -v?18:17
kgriffsmpanetta: sets logging level to INFO18:18
mpanettaoic18:18
malinikgriffs: do you need a log_path set ?18:18
mpanettaso verbose18:18
kgriffsmalini: yeah, I set log_path in my marconi.conf to /usr/local/var/log/marconi/queues.log18:18
kgriffsthen ran18:18
kgriffsmarconi-server -v18:18
kgriffsand got this: https://gist.github.com/anonymous/ff690b7888ca2bdc9eb918:18
kgriffsand nothing printed to console18:18
malinikgriffs: can you paste ur entire conf file?18:19
kgriffsoops18:19
kgriffscrap18:19
kgriffsI meant, I set log_file18:19
maliniaah ..ok18:19
kgriffsmal18:19
kgriffshttps://gist.github.com/anonymous/15a5d8e245fd1f8bd06018:19
kgriffsthere's my marconi.conf18:19
kgriffsalcabrera: can you review pls? https://review.openstack.org/#/c/76040/18:20
*** davidhadas has quit IRC18:21
*** davidhadas has joined #openstack-marconi18:22
malinikgriffs: This is my conf file http://paste.openstack.org/show/71839/18:23
maliniLet me try getting rid of verbose/debug18:23
mpanettaShouldnt use_syslog be false?18:23
mpanettaOr do the other things log to syslog?18:25
malinimpanetta: doesnt work either way with use_syslog18:26
mpanettaBlergh18:26
kgriffsfolks, please check your email if you are subscribed to open-stack dev18:29
kgriffsI have proposed flwang to be a core reviewer and need everyone's votes ASAP18:29
*** balajiiyer has quit IRC18:31
*** davidhadas has quit IRC18:31
malinionly if flwang solves the devstack issue ;)18:32
malinikgriffs: but I dont see any mail yet18:33
cpallareskgriffs, malini: what does ATC stand for?18:36
maliniActive Tech Contributor?18:37
cpallaresmalini: thanks18:39
kgriffscpallares: active technical contributor18:45
kgriffsmalini was 99% correct :)18:46
kgriffsmalini: sometimes the mailing list is a little slow18:46
kgriffsyou should see it now18:46
kgriffsmalini: I would just leave use_syslog = False18:48
kgriffsand make sure /var/log/marconi is accessible18:48
malinikgriffs: yeap & by default gate uses syslog false18:48
malinikgriffs: I tried with ur conf file & no longer see console o/p18:49
kgriffsmalini: is there a place that gates set up that is writeable by the devstack stuff?18:49
kgriffs(for logs?)18:49
malinikgriffs: yes18:49
maliniSee logs here http://logs.openstack.org/49/63449/18/experimental/check-tempest-dsvm-marconi/ce5cb55/logs/devstacklog.txt.gz from the last run18:49
malinisearch for iniset /etc/marconi/marconi.conf18:50
maliniI am making changes locally in my devstack server to try out stuff18:50
malinikgriffs: I tried with the exact same config you had & no longer see terminal o/p with 'marconi-server &'18:50
maliniBut devstack still doesnt like it18:51
maliniBut when I direct marconi-server o/p to afile in the devstack script, it still runs18:51
malinikgriffs: do u want to take a look at my devstack server?18:51
kgriffsso, looks like the other projects log under /opt/stack/....18:52
maliniyes..18:52
maliniBut I still cant figure out any difference between us & them18:52
kgriffsoh maybe not18:53
kgriffshmmm18:53
maliniI mean there is nothing the other projects specify in their devstack scripts18:53
kgriffsceilometer has /var/log/ceilometer-api18:53
malinionly ceilometer has a /avr/log specified like us18:53
kgriffsoic18:53
malinibut only ceilometer18:53
kgriffs"devstack doesn't like it"18:53
kgriffscan you point me to the error?18:54
kgriffsand is the error in the gate, or on your box18:54
maliniI can repro it in my box now18:54
maliniin a cloud server *18:54
kgriffsoh, that helps!18:54
maliniit seems like marconi crashes or hangs18:54
kgriffsat least we can iterate quicker18:54
maliniyes!18:54
kgriffsok, can we confirm it is writing the startup message to queues.log ?18:55
kgriffsrun with info level debug and we should see "Serving on host..."18:55
malinikgriffs: it does18:55
maliniI have a band-aid to get around this18:56
kgriffsdo we get the "terminating" message too18:56
malinihttps://github.com/openstack-dev/devstack/blob/master/lib/marconi#L15118:56
malinikgriffs: no terminating message18:56
kgriffsmalini: so, screen_it is the workaround?18:56
maliniif I change https://github.com/openstack-dev/devstack/blob/master/lib/marconi#L151 to screen_it marconi-server "marconi-server --config-file $MARCONI_CONF" >> some_random_file.log18:56
maliniit works18:56
kgriffsoic18:57
malinikgriffs: devstack runs without screen in gate..18:57
kgriffsand what is written to some_random_file.log when you do that?18:57
malinikgriffs: nothing (!)18:58
kgriffshmmm18:58
*** balajiiyer has joined #openstack-marconi18:58
kgriffstry this18:58
kgriffsin marconi/common/cli.py18:58
kgriffscomment out L3518:59
kgriffsalso, does screen_it itself require stdout or something?18:59
kgriffsoh, you said that doesn't run in gate19:00
kgriffswell, can we remove screen on your box too so we are identical with gate?19:00
malinikgriffs: I use use_screen=false in my devstack config19:01
maliniSo tht shud make it identical19:01
maliniBut I can totally remove it19:01
kgriffsoh, ok19:01
kgriffsbrb19:01
kgriffs(food)19:01
alcabrerame, too. I'm starving. :x19:02
*** alcabrera is now known as alcabrera|afk19:02
* flaper87 back19:11
flaper87holy moly, long backlog19:11
maliniflaper87: quick!! fix it!19:15
flaper87malini: what's the fix?19:16
maliniflaper87: thts for you to figure out ;)19:16
flaper87ahhahahahahaha19:16
maliniI have a band-aid, but no real fix19:16
*** alcabrera|afk is now known as alcabrera19:19
alcabreraGreen grapes are *really* yummy19:19
flaper87alcabrera: Yes they are :D19:19
kgriffsback19:19
kgriffsmalini: did commenting out L35 make any difference?19:20
malinikgriffs: no :(19:20
kgriffshmmm, ok19:20
kgriffsso we start up19:20
kgriffsnever see the termination message19:20
kgriffsmust mean crash or hang as you said earlier19:20
malinikgriffs: yes19:21
malini& re-directing stdout to a file seems to fix things19:21
kgriffsmore likely a hang19:21
kgriffssince a crash should still log the error19:21
kgriffs(given we catch all exceptions in cli.py)19:21
maliniif it hangs, I shud see marconi-server in ps19:21
kgriffswell, except for signals19:21
maliniI dont see the process running19:21
kgriffshmmm19:21
kgriffsthe plot thickens...19:21
kgriffswhen you do your workaround, you see it?19:22
maliniyes19:22
* kgriffs tears out hair19:22
* kgriffs what's left of it19:22
alcabreraD:19:22
malinimy hair is all gone after 3 weeks19:22
malinithis is dinner conv in my house these days..even my 1 year old knows19:23
kgriffsok, so it starts listening19:23
kgriffsthen disappears19:23
kgriffs...without a trace!19:24
* kgriffs queue scary music19:24
alcabrerait was the butler19:24
alcabrerathe butler killed marconi19:24
alcabrerahis name -19:24
alcabreraF. D. Redirect19:25
cpallareslol19:25
malinimpanetta tried adding strace &the last thing we saw was it trying to open a file descriptor19:25
kgriffsmalini: maybe we can try running it under supervisord19:25
malinialcabrera: :D19:25
kgriffsand see if supervisord logs what happens to the process19:25
alcabrera:)19:25
malinion it now..let me figure out how to do tht..19:25
kgriffsmalini: when you redirect output to a file is it 0 byte file?19:29
maliniyes19:29
kgriffsor maybe it contains a single null or newline?19:29
kgriffsoh darn, worth a shot19:30
malini-rw-rw-r-- 1 stack stack    0 Mar  3 18:48 m.log19:30
*** jmckind has quit IRC19:31
*** jmckind has joined #openstack-marconi19:32
*** balajiiyer has quit IRC19:32
kgriffsmalini: so, something needs stdout to be there19:34
kgriffseven though it doesn't actually write to it19:34
maliniyes19:35
kgriffsmpanetta, flaper87: know of a tool that will help us figure out what that "something" is?19:35
kgriffsmalini: I just got an evil thought19:35
kgriffsmaybe we just monkey-patch sys.stdout19:36
kgriffs(until we get to the root of this)19:36
kgriffsactually19:36
kgriffswait19:36
kgriffswe could monkey patch19:36
kgriffsand log when something tries to write to it!19:36
malinitht sounds like a good plan!19:37
kgriffsstand by19:37
malinikgriffs: with supervisord19:38
malini2014-03-03 19:37:24,869 INFO spawned: 'marconi-server' with pid 691219:38
malini2014-03-03 19:37:25,872 INFO success: marconi-server entered RUNNING state, process has stayed up for > than 1 seconds (startsecs)19:38
maliniI amtrying with debug now19:38
*** sballe has joined #openstack-marconi19:39
*** balajiiyer has joined #openstack-marconi19:39
malinikgriffs: http://paste.openstack.org/show/71855/19:40
*** sballe_ has quit IRC19:40
kgriffshmmm, looks like some FDs were opened for stderr and stdout?19:43
flaper87mmh, sorry, I'm in and out tonight!19:49
flaper87what have you guys figured so far?19:49
flaper87kgriffs: could you review the sql patches?19:49
kgriffsflaper87: will do in a few minutes - trying to get a stdout monkey patch to malini so she can try it19:51
kgriffsmalini: https://gist.github.com/anonymous/72982cfad1046d94462b20:01
malinikgriffs: shud I replace the cli.py with this ?20:01
kgriffsjust replace marconi/cmd/server.py20:01
kgriffs(server.py not cli.py)20:02
kgriffsthat should log an error message if something tries to write to stdout20:02
malinigreat !!20:03
alcabrerahehe20:03
alcabreramonkey patch the environment away20:03
alcabreraI hope to hear good news20:04
* kgriffs crosses fingers20:04
alcabreraoh yes20:04
alcabrerareviews20:04
alcabreralemme do that20:05
kgriffsalcabrera: http://paulmontgomerycode.wordpress.com/20:05
malinikgriffs: hmm..shud I have seen a message somewhere?20:05
kgriffspaulmo will be brain dumping in the next few weeks20:05
kgriffsyou two should sync up20:06
kgriffsmalini: only if something tried to write to stdout20:06
kgriffsfrom python land20:06
alcabrerakgriffs: I'll ping paulmo via email by the end of day. I've also been talking to Stefano M. about dropping barriers to entry for newcomers to OS20:07
kgriffssweet20:07
kgriffsmalini: no luck?20:07
malinino :(20:07
kgriffsmaybe also patch stderr20:07
kgriffssys.stdout = sys.stderr = MockSTDOut()20:07
*** sriram has joined #openstack-marconi20:09
kgriffsmalini: what does "screen_it" do?20:10
kgriffscould "screen_it" be trying to write out to the console itself?20:10
alcabrerakgriffs, flaper87: approved (https://review.openstack.org/#/c/76040/) test refactoring20:10
kgriffseven though it is disabled20:10
kgriffsalcabrera: w00t20:10
kgriffsthanks man20:10
kgriffsI'll pay you later. ;)20:10
alcabrerasweet. :)20:10
flaper87alcabrera: awesome!20:10
flaper87:D20:10
* alcabrera wants to rack up a high review combo20:11
*** jmckind has quit IRC20:12
openstackgerritA change was merged to openstack/marconi: test(wsgi): Refactor tests in preparation for API v1.1  https://review.openstack.org/7604020:12
malinikgriffs: let me chk20:12
kgriffsalcabrera: https://review.openstack.org/#/c/76480/420:12
malinikgriffs: but in tht case other projects shud have seen it too20:13
alcabrera+2'd, approved20:13
alcabrerasqlite is special20:13
kgriffsalcabrera: let's go through all the related patches20:13
alcabrerakk20:14
kgriffsflaper87: https://review.openstack.org/#/c/76481/20:14
flaper87kgriffs: yo20:14
kgriffsdo we have the work tracked somewhere to add a test?20:14
kgriffs"I think the bulk_delete deserves a more detailed one. However, since that affects all drivers, I'd prefer adding it later."20:15
mpanettaJust FYI, kgriffs, malini, all processes have FD's 0, 1, and 2 opened for them by the OS when it starts the process.  These of course are stdin, stdout, and stderr respectively.20:15
flaper87kgriffs: keep reading20:15
flaper87:P20:15
kgriffsah, found it20:16
kgriffsso many comments!20:16
flaper87yeah, and so confusing20:16
* flaper87 doesn't like gerrit's UI20:16
flaper87kgriffs: https://review.openstack.org/#/c/76483/ this one is tracked by the same bug on the previous review20:17
mpanettakgriffs: WRT your IoT talks... http://circuitcellar.com/wiznet2014/samplerequest/20:17
mpanettaI am trying to get in to that one.20:17
alcabreralots of merging coming soon20:19
kgriffsalcabrera: you missed one :D20:19
mpanettaI need a good project description... heh20:19
alcabreraworkers = 20 amused me. Serious hard work there.20:19
alcabreraIoT!20:19
alcabrerasuch a massive topic. :)20:19
mpanettaAll I know is I want to use Marconi heh20:20
malinikgriffs: this is where they do the log handling w/o screen https://github.com/openstack-dev/devstack/blob/master/functions-common#L96020:21
kgriffswoah20:22
mpanettahah20:22
kgriffsall kinds of redirects20:22
mpanettaNow you see!20:22
mpanettaWelcome to the rabbit hole kgriffs20:22
malinimpanetta: you can fix this!20:23
maliniI mean the devstack redirects ;)20:23
mpanettamalini: I think the correct term would be rewrite :P20:23
kgriffsok, so where are we on this?20:24
kgriffsthe stdout patch didn't help, or did it?20:24
mpanettastdout patch?20:25
malinikgriffs: it did not20:25
kgriffsmpanetta: I monkey-patched python's sys.stdout20:25
mpanettaTo do what?20:25
kgriffsmalini: one more thing20:25
mpanettadump everything?20:25
kgriffshttps://gist.github.com/anonymous/d450c94cb9929b8e1eec20:26
kgriffstry this version20:26
kgriffsit will just eat the messages20:26
maliniok20:26
kgriffs(rather than try to log them anywhere)20:26
mpanettaI still don;t understand why printing to stdout is so bad :(20:26
kgriffsmpanetta: IT_SHOULD_NOT_BE20:27
kgriffsso, when malini does an stdout redirect, we're gravy20:28
kgriffsafaik, '>>' just redirects stdout, not stderr20:28
mpanettayes20:28
alcabreraETOO_MUCH_DEVSTACK_WEIRDNESS20:28
kgriffsLOL20:28
mpanettaTo do stderr you would have to do >foo 2&>120:28
kgriffs"It worked on my box."20:29
mpanettaer20:29
mpanetta2>&120:29
kgriffsmalini: as a last resort can we do the >> trick and file a bug to continue working on finding the root cause?20:29
malinikgriffs: still no luck20:29
kgriffsdarn20:29
kgriffsok, I am officially out of ideas20:29
malinikgriffs: I am thinking of just submitting a devstack patch as a bandaid20:30
flaper87ah, my patch failed to merge20:30
flaper87I know it's kgriffs fault20:30
maliniI hope that it'll work in the gate20:30
kgriffsmalini: I'm OK with that as long as we file a bug to revisit20:30
maliniok20:30
kgriffsmalini: you may mention the bug in your commit msg20:30
mpanettaThey seemed ok with something like that, remember when I tried doing that patch?20:30
kgriffsat this point, it would seem that there are two possibilities20:31
kgriffsfirst, something is trying to use stdout before it hits our app20:31
kgriffsmaybe middleware?20:31
kgriffsno, that would still come in with server.py20:32
mpanettaI think it prints that lit is listening on port blah20:32
kgriffswell, something else20:32
maliniI'll open a bug & reference it in the patch20:32
maliniBefore that I'll make sure it really works on a brand new devstack server20:32
kgriffsOR something in our app is doing something with stdout other than writing, but still requires a console20:32
kgriffsmalini: kk20:32
maliniIt always seems to work here & fail at the gate20:32
kgriffsjoy20:32
kgriffsdarn20:33
mpanettaWhat else can you do to stdout besides write to it? heh20:33
kgriffsflaper87: looks like you have some rebasing to do! :(20:33
maliniI am going to write a primer on how NOT to do things , after this :|20:33
kgriffssrsly20:33
malinigot to go know..will submit my bug/patch later this evening20:33
mpanettaI think it would all work fine if it wasn't for all the wonky redirects20:34
kgriffsdevstack has failed it's mission if it is supposed to duplicate gate environment on your local box20:34
kgriffsmalini: ok20:34
kgriffsthanks!20:34
malinithank you kgriffs!!20:34
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Register sqlalchemy endpoints  https://review.openstack.org/7648220:34
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Don't raise `DoesNotExist` on claims deletion  https://review.openstack.org/7648320:34
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Enable pragma foreign key for sqlite  https://review.openstack.org/7648020:34
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Fix message's get and deletion  https://review.openstack.org/7648120:34
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Deprecate sqlite in favor of sqlalchemy  https://review.openstack.org/7648420:34
flaper87kgriffs: alcabrera there you go! :D20:35
kgriffsw00t20:35
alcabreralessee20:35
alcabreralots of patches20:35
alcabreragonna map (+2 <*> approve)' them20:35
alcabreranvm20:36
alcabrerakgriffs: is too much of a ninja20:36
alcabrera:P20:36
kgriffsonce all this is in, I'll rebase and address flaper87's comment on the bulk delete patch20:37
kgriffsthen it is off to the v1.1 races!20:37
kgriffsflaper87: btw, did you see my earlier comments about popping messages?20:37
flaper87kgriffs: nope, sorry, what was it about ?20:37
kgriffscome on man, you don't know about poppin'?20:38
kgriffshttp://www.youtube.com/watch?v=tc0t8Use3Wo20:38
kgriffssrsly, tho20:38
kgriffssomeone wants that so they can process things like loglines and op stats20:38
kgriffsthey are ok with the small risk of losing a message or two20:39
kgriffsso, they want bulk pop20:39
kgriffswhere, you can specify a limit of 10-5020:39
flaper87HOLYSHIT! that...dude....is...WTF?20:39
kgriffsyep. srsly.20:40
kgriffsthat guy is legit20:40
flaper87the holyshit was about the video anyway :P20:40
flaper87anyway, we've bulk_(get|delete|claim|post)20:41
flaper87having bulk pop just makes the API consistent so, why not?20:41
kgriffsw00t20:41
kgriffsbulk for everyone!20:41
mpanettabulk pop <tarts> ?20:42
alcabrerabulk eat20:42
kgriffsmmmm20:42
alcabreracustom HTTP verbs, ya know20:42
kgriffsI'm getting fat just thinking about that20:42
kgriffsstop it20:42
alcabreraHTTP EAT /v1/queues/poptarts20:42
mpanettaLOL20:42
alcabreralol20:42
mpanettaCan I do that in falcon yet? :P20:42
alcabreraI wonder...20:42
alcabreramaybe if defined an on_eat method for your responder20:43
alcabrera**if you defined...20:43
mpanettaWould it pick up on that? hmmm...20:43
alcabreraHMMM20:44
alcabrerahttp eat <url>20:44
alcabreraactually works!20:44
alcabreraI got this back from goo.gl20:44
mpanettaWoah20:44
alcabrera"The request method <code>EAT</code> is inappropriate for the URL <code>/</code>.  <ins>That’s all we know.</ins>"20:44
alcabrerahttpie <320:44
mpanettaOk so trying it now20:44
*** malini is now known as malini_afk20:44
kgriffsalcabrera: if we give jdpainter some bulk poptarts do you think he would help with our devstack issue?20:45
jay-atlHey Kurt, what is the process to fix the url not having /v1/ in the service catalog in OpenStack?20:45
mpanettaAww poo20:46
mpanettaI get bad request20:46
kgriffsalcabrera: even better, let's put them inside a brand new airplane20:46
kgriffswith his name on it20:46
alcabrerahahaha20:46
alcabrerasuch benefits20:46
alcabrerabrought to you by #openstack-marconi20:46
alcabrerayour own plane20:47
kgriffsjay-atl: i think we have to find the related project and submit a patch20:47
alcabrerabulk poptarts20:47
alcabrerayou name it20:47
alcabrerakgriffs delivers20:47
mpanettahaha20:47
alcabrerajay-atl, kgriffs: it's probably a keystone patch that needs sending20:47
kgriffsjay-atl: maybe ping dolphm in #openstack-dev ?20:48
kgriffshe is the Keystone PTL20:48
jay-atlOk, I'll try that. It's not something we have to do? It's them?20:48
kgriffsnot sure, but anybody can submit patches, so we can do one and they would review it20:48
kgriffsI just don't know exactly what needs to be patched TBH20:49
jay-atlOk, I'll ask him.. :-020:49
jay-atl:-)20:49
openstackgerritA change was merged to openstack/marconi: Enable pragma foreign key for sqlite  https://review.openstack.org/7648020:59
openstackgerritA change was merged to openstack/marconi: Fix message's get and deletion  https://review.openstack.org/7648120:59
kgriffshere we go!20:59
flaper87:D20:59
alcabreraw00t21:00
alcabreramy email is going to blow up21:00
alcabrerabut it's all for the greater good21:00
mpanettaalcabrera: Oh?21:02
alcabreraeverytime a patch gets rebased/approved/merged/ready-for-jenkins, I get a message21:02
openstackgerritA change was merged to openstack/marconi: Register sqlalchemy endpoints  https://review.openstack.org/7648221:03
alcabreraso in the next 10 minutes, I expect to go from 0 -> 6021:03
mpanettawhat repo is falcon under anyway?21:03
alcabrerarackerlabs/falcon, I believe21:03
mpanettahaha yeah that will be a few messages21:03
alcabreralemme double chekc that21:03
alcabrerahttps://github.com/racker/falcon/21:03
alcabrerait was racker21:03
mpanettaCool21:03
mpanettaThank you!21:04
alcabreranp. :)21:04
openstackgerritA change was merged to openstack/marconi: Don't raise `DoesNotExist` on claims deletion  https://review.openstack.org/7648321:04
mpanettaDoes any web standard actually support arbitrary verbs?21:04
alcabrerahmmm21:05
mpanettaThis whole thing has made me curious to how falcon works...21:05
alcabrerafor HTTP, I doubt it21:05
alcabreraI'm sure you can piggy-back anything at all on TCP21:05
alcabrerabrb21:05
*** alcabrera is now known as alcabrera|afk21:05
*** amitgandhi has quit IRC21:06
*** jamieh has quit IRC21:07
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Deprecate sqlite in favor of sqlalchemy  https://review.openstack.org/7648421:07
mpanettaApparently HTTP does allow custom methods...21:20
kgriffsyes, but that's EVIL21:20
* kgriffs ducks21:20
mpanettaEVIL is fun :P21:20
kgriffsum, I guess if someone really wants custom methods in falcon, they can send me a patch21:20
mpanettalol21:21
kgriffshttp EVIL http://example.com/science21:21
mpanettajoyent supports them :P21:21
mpanettakgriffs: How is that evil?21:21
kgriffsoh, I'm just kidding. Kind of.21:22
mpanettaDo I need some learning? heh21:22
kgriffswhen you support custom verbs, it is harder to use your service from standard tools/libs21:22
kgriffsthat's all21:22
mpanettaOh, no I meant the example.com thing21:22
kgriffsI personally wouldn't design an API with them21:23
kgriffsmpanetta: haha21:23
kgriffsoic21:23
kgriffsscience isn't evil21:23
kgriffsbut if you wanted to *do* evil science21:23
kgriffsthat would be one way21:23
kgriffsi mean21:24
mpanettaUnless it was for something silly21:24
kgriffsuse science for an evil goal21:24
mpanettaAh that is done all the time, unfortunately21:24
kgriffsyou know, like to create dark chocolate that is 110% cacao21:24
mpanettaHaha21:25
mpanettaNow that *IS* evil!21:25
mpanettaAll I wanted was to be able to write a falcon app that could eat pop tarts.  Is that so bad? :P21:26
kgriffslol21:26
kgriffsno, I suppose not21:26
kgriffsit would make for a nice demo, anyway21:26
*** sriram has left #openstack-marconi21:26
kgriffshttp TOAST http://iot.example.com/poptart21:26
kgriffsmpanetta, cpallares: if someone were to submit an elegant patch to falcon for supporting custom verbs, I may be convinced to merge it21:28
kgriffsjust sayin'21:28
kgriffsespecially if it came with a hilarious unit test21:28
mpanettalol21:29
mpanettaaww iot.example.com does not work?21:30
mpanettapoptarts and tacos21:31
kgriffsflaper87: I think that patch is still failing21:32
cpallareskgriffs: I might conspire to add a poptarts one with mpanetta.21:33
cpallareshilarious unit test included.21:33
cpallares:P21:33
kgriffsrock on21:33
mpanettahaha21:34
openstackgerritKurt Griffiths proposed a change to openstack/marconi: fix(MongoDB): Driver does not retry on AutoReconnect errors  https://review.openstack.org/7516521:34
*** alcabrera|afk is now known as alcabrera21:35
alcabreraI'd be surprised if falcon didn't already support custom HTTP verbs21:35
alcabreraI don't remember us explicitly disallowing them21:35
kgriffsalcabrera: I think it does as a side-effect of the way generic responders work.21:36
kgriffsbut, I could be wrong21:36
alcabreramy thoughts exactly, kgriffs21:36
kgriffsalcabrera, cpallares: your mission, should you choose to accept it, is to "persuade" mpanetta to submit his first Falcon patch.21:36
cpallareskgriffs: hahaha21:37
alcabreramission started, kgriffs. I sent mpanetta *the link*21:37
alcabrera:D21:37
mpanettaBTW, either way it needs to be fixed, currently if you send an unsupported method it returns http 400, which should apparently be a 405 or 501 according to what I have read..  But maybe not :P21:37
mpanettahaha YEah I am already looking at the link ;)21:37
cpallareswoo alcabrera, you're ahead of the game.21:37
mpanettaFound the Requests Class21:38
* mpanetta is insane :P21:38
mpanettahttp CATS http://cat.me21:39
mpanettaheh21:39
alcabreralol21:39
alcabrerahttp PURRFECT http://cat.ly21:39
alcabreraI'm out for the night.21:40
alcabrerasee you all tomorrow.21:40
mpanettaHave a good night!21:40
alcabreramore patches, more poptarts21:40
alcabrerao/21:40
*** alcabrera has quit IRC21:41
kgriffsneed some love here: https://review.openstack.org/#/c/76678/21:42
kgriffssee my response to flaper8721:42
kgriffsflaper87: jenkins is gunnin' for ya. https://review.openstack.org/#/c/76484/21:43
openstackgerritKurt Griffiths proposed a change to openstack/marconi: feat(api): API v1.1 bulk delete change  https://review.openstack.org/7609321:52
kgriffscpallares: can you sanity-check this patch for me?21:53
kgriffshttps://review.openstack.org/#/c/76678/21:53
cpallareskgriffs: sure thing21:53
*** amitgandhi has joined #openstack-marconi21:53
kgriffsthanks!22:00
flwangkgriffs: ping22:09
*** ayoung is now known as ayoung_afk22:09
*** amitgandhi has quit IRC22:10
*** amitgandhi has joined #openstack-marconi22:11
kgriffsflwang: pong22:14
flwangkgriffs: may I know what's your original thought about this https://bugs.launchpad.net/marconi/+bug/1222932 ?22:15
flwangkgriffs: btw, thanks for the nomination, I will definitely contribute more for the cure project22:16
kgriffs:)22:16
flwangkgriffs: I found a lot of fun at here :)22:16
kgriffsglad to hear it!22:16
kgriffsif I remember correctly (this is an old bug) there were some functions that were creating functions for filtering lists and stuff22:16
kgriffsmaybe filter isn't the right term22:17
kgriffsmore like "transform"22:17
kgriffssort of like named lambdas22:17
kgriffsdoes that make sense?22:17
flwangkgriffs: embedded functions, right?22:17
kgriffsright22:17
kgriffsdef foo(...):22:17
flwangbut we'd like to see make them like inline/lambda, exactly?22:17
kgriffs    def bar():22:17
kgriffs        pass22:18
kgriffs    [bar(x) for x in y]22:18
kgriffsstuff like that22:18
flwangcool, I can see your point22:18
kgriffsflwang: we'd like to make them "static"22:18
kgriffsso, if there is a nice way to declare the function once rather than making the interpreter recreate the function each time22:18
flwangkgriffs: yep, that would be better22:18
kgriffsthat reminds me, I need to take a look at that /health patch22:19
flwangkgriffs: haha, yep, that's my next request :D22:19
kgriffs;D22:19
flwangkgriffs: could you revisit my patch? about the overall shape22:19
kgriffskk22:19
kgriffswill do!22:20
flwangkgriffs: I will add test cases after your v1.1 patch landed22:20
kgriffsflwang: the test refactoring landed just a little while ago22:20
kgriffsso you should be all set22:20
flwangkgriffs: cool, thanks. btw, do you think we're safe given the FF is coming...22:20
flwangkgriffs: nice to hear ;)22:20
kgriffsI am working hard on v1.1 stuff22:21
kgriffsif we don't get everything done by EOD tomorrow22:21
flwangkgriffs: got it, count me22:21
kgriffsthen we will just have to go to the ATL summit and say that v1.1 stuff is experimental22:21
flwangkgriffs: oh, nope22:21
kgriffsby that time it should all have landed in the mainline, just won't be in the tarball for icehouse22:21
kgriffsI'm a hopeless optimist, so I am going to work on it down to the wire. :)22:22
kgriffsotherwise, I think the other bp's are on track22:22
flwangkgriffs: got it22:22
kgriffsrock on22:22
flwangkgriffs: coding...22:23
*** flaper87 is now known as flaper87|afk22:30
*** Kuwagata has quit IRC22:43
*** mpanetta has quit IRC22:47
*** balajiiyer has quit IRC23:02
*** mwagner_lap has joined #openstack-marconi23:02
*** jergerber has quit IRC23:02
*** tedross has quit IRC23:04
*** balajiiyer has joined #openstack-marconi23:04
kgriffssee you later folks23:12
kgriffsI should be back online in a few hours to assist with any i-3 stuff23:12
*** balajiiyer has quit IRC23:12
*** amitgandhi has quit IRC23:14
kgriffsflwang: feel free to assist with v1.1 bp's23:15
kgriffshttps://launchpad.net/marconi/+milestone/icehouse-323:15
kgriffssame goes for anyone else: ^^^23:15
kgriffscpallares: ^^^23:16
kgriffsttfn23:16
*** kgriffs is now known as kgriffs_afk23:26
« Sunday, 2014-03-02 Index Tuesday, 2014-03-04 »

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