Thursday, 2014-07-31

*** flwang__ has joined #openstack-marconi00:01
*** flwang__ has quit IRC00:06
*** malini has joined #openstack-marconi00:23
*** rwsu has quit IRC00:31
*** malini has quit IRC00:58
wpfvkmc:  ping01:11
vkmcwpf, pong01:12
wpfvkmc:  new name of Zaqar?01:12
wpfwhat does it mean , -:)01:13
vkmcwpf, yeah... earlier today we decided to change Naav for Zaqar01:13
vkmcit was Zaqar or TamTam01:13
vkmcit means..01:13
wpfah, got it ,01:14
vkmcdo you like it or do you prefer Naav?01:15
wpfI remembered flwang told me that the Naav is the the river name somewhere01:16
wpfand from the Wiki,  Zaqar is the messenger,  so it should be more meaningful01:16
wpfI am just curious that how you find the name, go through the dictionary and oh this one is ok ?  ^^01:18
vkmcwe had a couple of meetings to decide the new name for the project01:22
vkmcwe brainstormed in an etherpad01:22
vkmcand well, lots of things came up there... under the effects of coffee and gummybears01:23
wpfyes, I can imagine that01:23
vkmcI'm heading off for dinner o/01:24
wpfgo ahead01:25
wpfflwang:  ping01:30
flwangwpf: yes01:31
wpfis there a tape for the IRC chat history ?  , I want to go through01:31
wpfnot sure how so say it , I just want to go through  the chat log of some day01:32
flwangfor meeting channel, I think there is, but as for the other channels, I don't know01:33
wpfOk, got it ,seems that the IRC client can write them into the log file01:35
*** nosnos has joined #openstack-marconi01:50
*** flwang__ has joined #openstack-marconi02:02
*** flwang__ has quit IRC02:07
*** oz_akan_ has joined #openstack-marconi02:11
*** oz_akan has quit IRC02:11
*** vkmc has quit IRC02:28
*** prashanthr_ has quit IRC02:30
*** haomaiw__ has quit IRC02:33
*** haomaiwang has joined #openstack-marconi02:34
*** haomai___ has joined #openstack-marconi02:50
*** haomaiwang has quit IRC02:52
*** haomaiwang has joined #openstack-marconi03:00
*** haomaiwang has quit IRC03:03
*** haomai___ has quit IRC03:03
*** haomaiw__ has joined #openstack-marconi03:04
*** haomaiw__ has quit IRC03:11
*** haomaiwang has joined #openstack-marconi03:12
*** haomai___ has joined #openstack-marconi03:14
*** haomaiwang has quit IRC03:17
*** oz_akan_ has quit IRC03:19
*** prashanthr_ has joined #openstack-marconi03:36
*** haomai___ has quit IRC03:44
*** haomaiwang has joined #openstack-marconi03:44
*** prashanthr_ has quit IRC03:46
*** haomaiwang has quit IRC03:47
*** nosnos has quit IRC03:52
*** boris-42 has quit IRC03:57
*** flwang__ has joined #openstack-marconi04:03
*** oz_akan has joined #openstack-marconi04:05
*** boris-42 has joined #openstack-marconi04:07
*** flwang__ has quit IRC04:08
*** prashanthr_ has joined #openstack-marconi04:27
*** Rajalakshmi has quit IRC04:36
*** oz_akan has quit IRC04:42
*** oz_akan has joined #openstack-marconi04:43
*** chandankumar has joined #openstack-marconi04:45
prashanthr_here is the daily log of the irc chat history04:45
prashanthr_i guess this is what you are looking for04:45
wpfprashanthr_:  exactly !  Thanks04:45
prashanthr_welcome :)04:46
*** oz_akan has quit IRC04:47
*** nosnos has joined #openstack-marconi04:50
prashanthr_flwang, wpf: Can you please review whenever you are free ?04:52
wpfI will try , ^^05:00
flwangprashanthr_: yes, it's on my list05:02
flwangand we will make it happen in J-305:02
prashanthr_flwang: thank you ! :) Sure.05:03
prashanthr_when's the deadline for J-3 ?05:03
*** cpallares has quit IRC05:28
*** oz_akan has joined #openstack-marconi05:44
*** oz_akan_ has joined #openstack-marconi05:45
*** oz_akan has quit IRC05:45
*** oz_akan_ has quit IRC05:50
*** k4n0 has joined #openstack-marconi05:51
*** flwang__ has joined #openstack-marconi06:04
*** flaper87|afk is now known as flaper8706:05
*** flwang__ has quit IRC06:09
*** prashanthr_ has quit IRC06:20
*** prashanthr_ has joined #openstack-marconi06:22
flaper87flwang: around?06:27
flaper87flwang: are you working on the sqla thing?06:27
flwangflaper87: o/06:29
flwangflaper87: yes, if you want to pick, i'm ok06:29
flaper87flwang: if you're working on it now, then that's fine.06:30
flaper87we'd like to get those done asap06:30
flaper87I'm asking because you mentioned you don't have enough bandwith06:30
flwangflaper87: ah, I mean I don't have TOO much bandwidth06:31
flaper87flwang: as in, you're not busy ?06:31
flwangflaper87: but I assume it's a small change. maybe just add a check, to see if the uri scheme is in the expected list06:31
flwangflaper87: oh, come on06:32
* flaper87 is messing w/ flwang06:32
flwangflaper87: I have to take care a lot of different stuff, since now the team is very small06:32
flwangI'm not the guy working at IBM06:32
flaper87flwang: so, we'd like to get it done today/tomorrow06:32
flaper87any chance you can get it done?06:32
flaper87it's a matter of removing un-used code06:33
flwangI can submit a  patch today06:33
flaper87flwang: the sqlalchemy data driver should go away06:33
flaper87we'd like to keep *just* the control-api06:33
flwangI see06:34
flaper87flwang: so, can you do it?06:35
flwangI will try to submit a patch today, but I can't grantee it06:35
* flaper87 has freak mode on06:35
flaper87flwang: in that case, would you mind if I take it?06:35
flaper87I can work on that now06:35
flwangflaper87: fine06:35
flaper87flwang: coooooool06:36
flaper87I'll review your health patch asap06:36
flwangok, thanks06:36
wpfprashanthr_:  ping06:37
prashanthr_wpf: Hello06:43
prashanthr_I just slept off on my desk :)06:43
wpfI'd review your PS, and sorry for -106:44
wpfI am not familiar with Redis, I believe it works, anyway, there are some comments , ^^06:46
flaper87wpf: don't be sorry about -1, authors should thank for -1 :)06:46
*** oz_akan has joined #openstack-marconi06:46
prashanthr_wpf: Yes as flaper87 says happy for -1's06:46
prashanthr_thanks a lot for reviewing bcaz more reviews -> lesser bugs and faster check in's06:47
* prashanthr_ Will be happy to see his code in the repo :)06:47
prashanthr_and of course more bugging of flaper87 ;)06:47
flaper87wait what?06:47
flaper87I'm not even here06:47
prashanthr_he he more reviews-> more bugging of flaper87 ;)06:48
flaper87prashanthr_: who are you talking to? Who's that flaper87 ?06:48
*** prashanthr_ is now known as flaper87_06:49
flaper87_that's me06:49
*** flaper87_ is now known as prashanthr_06:49
*** oz_akan has quit IRC06:51
wpfa newbie question,  I always saw there are some 'actions' or 'emotion' you do in the IRC,  is there a guide for such things?06:52
wpfwhat I am saying are those lines  start with a '*' in the left side06:55
*** reed has joined #openstack-marconi06:57
*** reed has quit IRC06:57
prashanthr_wpf: yes06:58
prashanthr_you can use the "/me : message"06:58
prashanthr_then you can see the expressions/opinions kind of messages we post06:58
prashanthr_try it now06:58
* wpf learn new skills from prashanthr_ 06:59
prashanthr_wpf: Yaaaaaay :)06:59
wpfit works06:59
prashanthr_more of such commands here06:59
wpfgreat, Thanks prashanthr_06:59
prashanthr_no problems06:59
prashanthr_btw great catch about the _exists function06:59
prashanthr_saves a real o(n) operation07:00
prashanthr_which will be a real pain when the size of the queue grows :)07:00
wpfhmm , I forgot what I found .....07:00
* prashanthr_ Applauds wpf :) 07:00
prashanthr_wpf: It is about checking if a message exists before raising an exception or fetching the message07:03
prashanthr_which could be done by fetching the message directly :)07:03
wpfyes,  I remembered it at last07:03
prashanthr_wpf: can you explain to me this comment07:07
prashanthr_the second one in the file07:08
wpfdo  we need the () for the 'help' value?07:08
wpfto concatenate the mulit-line string literal?07:09
prashanthr_aahh yes we do.07:10
wpfthat's fine07:11
prashanthr_wpf: Can u explain to me how to use marconi.i18n in place of from marconi.openstack.common.gettextutils import _07:17
wpffrom marconi.i18n import _07:18
prashanthr_that's all is the change ?07:18
wpfwe use the oslo.i18n now07:19
wpfand marconi.i18n wrap it07:20
prashanthr_great :)07:24
wpfif you are working on the next comment on 'msg_expired_filter', please notice that I just add another comment for it , you can refresh it , -:)07:26
prashanthr_wpf: Saw it :)07:28
prashanthr_I have given all the reviewers enough nightmares with my documentation07:29
prashanthr_and NOTE's :P07:29
prashanthr_especially kgriffs|afk and vkmc07:29
openstackgerritPrashanth Raghu proposed a change to openstack/marconi: Implements Queue,Message and Claims Controllers for Redis
wpfso fast!07:39
prashanthr_wpf: Ha ha yes :) I actually finished my thesis work requirement for my master's program07:40
prashanthr_so totally free now07:41
wpfcomputer science?07:42
prashanthr_wpf: thank u :) Just waiting for my presentation dates now.07:43
*** AAzza_afk is now known as AAzza07:43
wpfthose in China already graduate from college, :P07:45
*** oz_akan has joined #openstack-marconi07:47
prashanthr_wpf: Sorry did not get you07:48
*** prashanthr_ has quit IRC07:48
*** prashanthr_ has joined #openstack-marconi07:50
*** oz_akan has quit IRC07:51
wpfI mean that colleges in China already graduated about one month ago07:52
wpfand what's your major?07:53
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Remove sqlalchemy's DataDriver
prashanthr_wpf: Actually I was supposed to graduate as well08:02
prashanthr_delayed due an independent thesis work i chose08:02
prashanthr_my major is "Wireless Sensors and Internet of things"08:03
*** flwang__ has joined #openstack-marconi08:05
*** prashanthr_ has quit IRC08:09
*** flwang__ has quit IRC08:10
*** prashanthr_ has joined #openstack-marconi08:11
flaper87AAzza: there?08:13
flaper87you may want to take a look here:
flaper87(especially where your nickname is :P)08:13
*** prashanthr_ has quit IRC08:21
*** nosnos has quit IRC08:26
*** nosnos has joined #openstack-marconi08:26
*** ykaplan has joined #openstack-marconi08:28
*** flwang__ has joined #openstack-marconi08:36
*** oz_akan has joined #openstack-marconi08:48
*** oz_akan has quit IRC08:52
AAzzaflaper87:  yeap, looking.08:55
wpfflaper87:   I saw you remove the sqlalchemy related code , but we still default the sqlite when marconi starting , will that be affected?09:05
*** malini|afk has quit IRC09:12
*** malini|afk has joined #openstack-marconi09:12
*** malini|afk is now known as malini09:13
flaper87wpf: still working on it, I think we'll need a better road for this09:13
*** prashanthr_ has joined #openstack-marconi09:20
prashanthr_flaper87: Jenkins is failing for your patch09:25
flaper87yeah I know09:25
flaper87considering another approach09:25
flaper87we need to follow a deprecation path, we can't just remove it all together09:26
prashanthr_okay was just curious because most cases were failed so thought there must be a problem with jenkins09:26
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Mark sqlalchemy's data driver as deprecated
flaper87prashanthr_: wpf ^^^^ This one09:29
prashanthr_flaper87: okay :)09:31
*** flaper87 is now known as flaper87|afk09:38
*** prashanthr_ has quit IRC09:44
*** prashanthr_ has joined #openstack-marconi09:45
*** oz_akan has joined #openstack-marconi09:48
*** oz_akan has quit IRC09:53
*** prashanthr_ has quit IRC10:01
*** ykaplan has quit IRC10:29
*** oz_akan has joined #openstack-marconi10:49
*** oz_akan has quit IRC10:53
*** flaper87|afk is now known as flaper8711:36
*** ykaplan has joined #openstack-marconi11:40
*** vkmc has joined #openstack-marconi11:45
*** vkmc has quit IRC11:45
*** vkmc has joined #openstack-marconi11:45
vkmcgood morning!11:47
flaper87vkmc: morning11:50
flaper87why can't ppl have decent opened wireless spots?11:50
flaper87They're all slow... I mean, if you're going to leave it open, you should at least make sure it's fast enough11:51
*** abettadapur has joined #openstack-marconi11:51
vkmcI hear you11:52
*** k4n0 has quit IRC12:14
vkmcI'm going through Marconi docs to find the right place to put the marconi.conf guide12:23
flaper87vkmc: holy shit, how did you do that?12:23
flaper87I was exactly writing that12:24
vkmcwhere do we host the user guide? or we don't?12:24
vkmcthis are the developers docs
flaper87I wanted to know whether you could tackle those now or later12:24
vkmcyup I'm tacking that now :p12:24
flaper87awesome, thanks12:24
flaper87yeah, iuugh, those docs need love12:25
flaper87vkmc: best thing to do is contact anne12:25
vkmcflaper87, already did :)12:25
flaper87pls, get back with your findings :D12:25
vkmcI'm waiting for her reply12:25
flaper87vkmc: have I told you how amazing you are?12:25
vkmcaww, you are amazing too flaper8712:26
vkmcanyhow, I'm a bit worried that the developer docs have things more related to user than the developer12:27
*** flwang__ has quit IRC12:27
vkmcI hope Anne can give me some information about what they expect the developer guide and user guide have to graduate12:28
flaper87yeah, I'm afraid that's the case :(12:28
*** flwang__ has joined #openstack-marconi12:28
flaper87let me know if we need to organize an all-hands doc day12:28
vkmcsure, I'll let you know later today12:28
vkmcI'm checking out the docs wiki, it's a good starting point
*** flwang__ has quit IRC12:32
*** jasondotstar has joined #openstack-marconi12:36
*** nosnos has quit IRC12:40
*** jasondotstar has quit IRC12:42
*** sriram has joined #openstack-marconi12:42
*** jasondotstar has joined #openstack-marconi12:56
*** mwagner_lap has joined #openstack-marconi12:59
*** oz_akan has joined #openstack-marconi13:01
*** mpanetta has joined #openstack-marconi13:08
*** catherin_ has joined #openstack-marconi13:08
*** balajiiyer has joined #openstack-marconi13:10
vkmcwell it looks like the docs are a blocker for graduation13:13
vkmcwe have to put those in order13:13
*** mwagner_lap has quit IRC13:15
*** jay-atl has quit IRC13:15
flaper87vkmc: oh yeah, they are. How bad is our status?13:15
*** chandankumar has quit IRC13:15
*** mwagner_lap has joined #openstack-marconi13:15
*** jay-atl has joined #openstack-marconi13:15
vkmcwe are not that bad because most of the docs are in the wiki13:16
*** chandankumar has joined #openstack-marconi13:16
vkmcfor the user we require docs for the API use only (we don't have a CLI and integration with the Dashboard yet)13:17
flaper87vkmc: gtk, how much work do you think we have left to do? Is that something we can estimate?13:17
flaper87(in 5)13:17
flaper87(in %)13:17
vkmcinstallation and configuration docs, we have them13:17
vkmcwell, in addition to add the configuration files docs we have to migrate some docs in the wiki to the user guide13:18
*** balajiiyer has left #openstack-marconi13:18
vkmcis hard to estimate13:18
vkmcbut I'm optimistic13:18
* vkmc avoids saying numbers :p13:21
flaper87ok, lets put it this way. Is that something we can tackle in a 1 all-hands documentation day ?13:24
flaper87doc-squashing day13:24
flaper87although squashing is not exactly what we want to do13:24
*** malini1 has joined #openstack-marconi13:25
vkmcflaper87, it is yes13:26
flaper87ok, cool13:26
flaper87lets schedule that next week13:26
flaper87vkmc: since you've dug into this, do you want to send the email?13:27
flaper87I can send it if you prefer13:27
vkmcflaper87, I can do that, I'll add some day alternatives13:28
flaper87vkmc: awesome13:28
flaper87lets gather more info and send that email today13:28
vkmcwill do that13:29
flaper87thanks, thanks, thanks thanks13:29
vkmcI'll stub out what needs to be done so we can divide work13:29
vkmchola malini! :)13:29
flaper87malini1: hola caracola13:30
malini1ok..what is caracola13:31
mpanettaSounds like a drink :P13:31
flaper87oh my dear lord, I broke everything13:32
mpanettaflaper87: What ya go and do that for?13:32
flaper87malini1: caracola is snail but that greeting is common in spanish13:33
* flaper87 is not saying malini1 is a snail13:33
flaper87(just in case)13:33
flaper87mpanetta: I DON'T KNOW13:33
flaper87I think I'm just dumb13:34
malini1It might not be a bad thing to a snail13:34
mpanettaflaper87: You?  Dumb?  No way man.13:34
malini1As long as I don't end up in somebody's pasta, i.e.13:34
*** jasondotstar has quit IRC13:35
mpanettaOh pretty:
*** jmckind has joined #openstack-marconi13:36
vkmcin some latin american countries people eat those :x13:36
flaper87not just latin-america, trust me13:36
flaper87I... would never...13:36
mpanettaHey, you never know, they might be tasty!13:37
*** catherin_ has left #openstack-marconi13:37
* vkmc shivers13:37
mpanettaflaper87: Damnit, you and your beautiful pics of the lake13:37
flaper87mpanetta: people say they are13:37
vkmcnope nope nope13:37
flaper87I don't want to know13:37
malini1vkmc: its just a matter of what you are conditioned to think of as food13:37
flaper87mpanetta: for you it's a pic, for me it's my current view13:37
mpanettaLooks so much like Lugano...13:37
* flaper87 wants to make mpanetta more jealous13:38
mpanettaDamn you :P13:38
malini1thanks flaper87! I have a ugly brown door right infront of my eyes13:38
vkmcmalini1, yeah, I guess it's a matter of getting used to it13:38
flaper87malini1: you could put the picture I just shared there13:39
mpanettahaha too bad we don't have an inkjet plotter in the office.13:39
malini1flaper87: so it can remind what a wonderful view you have?13:39
flaper87malini1: exactly13:41
flaper87this will help reminding you how good the coffee is here:
flaper87one problem, I'm running out of battery13:41
flaper87I'll have to go back to my cave13:42
flaper87or sit in a roof-top bar ;)13:42
mpanettaWhy must you tease us so mercilessly flaper87?13:43
flaper87because I want the best for you all13:43
flaper87ok, that didn't help,right?13:43
flaper87damn.. because I love you all ?13:43
mpanettaOk, so you will send us plane tickets to your home? :P13:43
sriramjust read all the meeting logs…13:43
flaper87mmh, ok ok, what about this one: Because one day, we'll do a mid-cycle meetup here13:43
* sriram phew13:43
malini1sriram: our new leader is boosting the team morale now13:44
sriramflaper87: \m/13:44
srirambtw good morning all :)13:45
mpanettaI wonder how much plane tix to italy are :P13:45
mpanettamorning sriram :)13:45
*** prashanthr_ has joined #openstack-marconi13:51
prashanthr_Good morning all :)13:52
mpanettamorning prashanthr_13:53
prashanthr_mpanetta: Hello :)13:53
AAzzahi, all morning people13:56
flaper87AAzza: 'sup girl ?13:59
flaper87how ya doing?13:59
flaper87did you go through the etherpad?13:59
AAzzaflaper87: yeap, trying to run benchmarking tool in loop and produce some results.14:00
AAzzaflaper87: for now thinking about storing all results to file and drawing plots with gnuplot as it is easiest way to start14:02
*** amitgandhi has joined #openstack-marconi14:02
AAzzaflaper87: not sure in what way it should be done, but will start with `anything working` and then will discuss together14:04
AAzzaflaper87: it is hard to kill perfectionist inside, that do not like to do `anything working`:-(14:05
*** chandankumar has quit IRC14:05
openstackgerritVictoria Martínez de la Cruz proposed a change to openstack/python-marconiclient: Adds extra info about python-marconiclient and usage instructions
AAzzabtw, other openstack project afaik don't use netiher gnuplot nor matplotlib, all i saw is generation of html page that draws plots using js14:07
AAzzabut we should start with smth and first agree on what and how should be plotted)14:08
AAzzaeither i'm talking to myself or inet is broken14:08
malini1AAzza: Rally folks might be already doing this14:09
*** AAzza1 has joined #openstack-marconi14:10
openstackgerritVictoria Martínez de la Cruz proposed a change to openstack/python-marconiclient: Adds extra info about python-marconiclient and usage instructions
*** AAzza1 has quit IRC14:15
AAzzamalini1: yes, they are doing (and it is they who use mako templates -> html +js for drawing), but i'm not sure why we decided to write our tool and thus don't know what and how to use.14:19
AAzzalast task is was to write simple working piece of code)14:19
malini1AAzza: The reason we started writing our own was because rally didnt support distributed loadgens etc - need to generate the volume of traffic we need14:20
malini1But the intent all along has been to merge this with rally whenever hey are ready14:20
prashanthr_AAzza: Do we need to do anything to enable the redis tests in gate ?14:23
AAzzamalini1: thanks for explanation, will look on their approach and maybe try to use similar formats\ideas to be at most compatible14:23
AAzzaprashanthr_:  not sure what to do in particular, but know that mongodb tests are disabled, and it looks that now redis also, need to ask someone14:24
*** oz_akan has quit IRC14:25
prashanthr_i have really shallow knowledge on the testing code base of marconi14:25
AAzzamalini1: maybe you know? redis tests for prashanthr_ driver is turning on the same ways as mongodb - via env variable. how to make them run on the gate?14:26
*** oz_akan has joined #openstack-marconi14:27
*** devanand1 is now known as devananda14:27
malini1AAzza: Does redis come default in the gating linux images?14:27
malini1Trusty/ Precise/ CentOS614:28
malini1Or does it have to be installed?14:28
*** flwang__ has joined #openstack-marconi14:28
AAzzamalini1: should be installed afaik14:29
*** kgriffs|afk is now known as kgriffs14:29
*** devananda has left #openstack-marconi14:32
*** flwang__ has quit IRC14:33
kgriffssriram: ping14:35
*** vkmc has quit IRC14:41
*** rwsu has joined #openstack-marconi14:42
kgriffsflaper87: btw, the 1:1 incubated project sync today with ttx has been canceled.14:44
*** tonytan4ever has joined #openstack-marconi14:45
kgriffssriram: I need to find out the status on benchmarking hardware14:45
*** tonytan4ever has quit IRC14:46
malini1AAzza: sorry had to step away for a meeting14:49
malini1AAzza: if it is available in the default image with the right version you need, it is straight forward14:50
malini1If it does not come with the default image, we might have a problem14:50
malini1Infra does not like to install anything on top of default images, except python packages14:51
malini1Try pinging #openstack-infra in tht case14:51
AAzzamalini1: i believe it is available  and what to do if everything is ok?14:55
*** cpallares has joined #openstack-marconi14:55
kgriffsAAzza: looks like prashanthr_ put an env variable flag to enable redis tests14:56
*** tonytan4ever has joined #openstack-marconi14:57
AAzzakgriffs: yeah, flag)14:57
kgriffsI think once we have mongo and redis available in the gate with the appropriate permissions, we should just remove those flags and devs will need to run localhost redis and mongo when testing their own patches. test all the things all the time. :p14:57
kgriffsthat's my $0.02 anyway14:57
kgriffswe can keep the flags until Kilo and remove in k-114:58
kgriffshave to see what flaper87 thinks, but that's my suggestion14:58
AAzzakgriffs: totally agree about that)  it is always sad, when run local tests with mongo enaibled and find out that it is broken for sometime:)15:00
kgriffssad indeed. :(15:00
sriramkgriffs: pong15:01
*** flaper87 is now known as flaper87|afk15:01
kgriffssriram: o/15:01
*** prashanthr_ has quit IRC15:02
kgriffsAAzza: i made a related bug for future reference.
kgriffssriram: did you see my email?15:02
sriramThe webheads have two vcpus, 4gb ram, 160gb disk, and 400meg network.15:02
sriramkgriffs: no15:02
sriramlet me check15:02
kgriffssriram: so they are virtual, not bare metal?15:03
sriramnot bare metal15:03
kgriffssriram: how many boxes total, and what roles?15:04
sriram4 webheads, 1 ha proxy load balancer 1 mongo replica set15:04
*** prashanthr_ has joined #openstack-marconi15:04
sriramone sec, let me gather all the info15:04
*** whenry has joined #openstack-marconi15:05
sriramcatalog replica set(3)is similar, the mongo replica set(3) is 4vcpus, 8gig ram, 320gig disk and 600meg network15:07
sriramthe haproxy instance is two vcpus, 4gb ram, 160gb disk, and 400meg network.15:08
sriramin total 1 + 4 + 3 + 3, 11 boxes15:08
sriramkgriffs: ^15:08
sriramI added performance tweaks, from your gist.15:10
kgriffsok, that sounds good. Would you mind making a rough diagram of that?
sriramsure, give me a moment15:12
*** AAzza is now known as AAzza_afk15:12
kgriffsyou can export to png and upload to
kgriffsthat would be awesome15:12
sriramwill do kgriffs15:12
prashanthr_Hi, I just need a small help15:13
prashanthr_pyredis stores a dictionary into the db15:13
prashanthr_as a string15:13
*** ykaplan has quit IRC15:13
prashanthr_and when it is retrieved back it is a python string15:13
prashanthr_so i planned to conert it into a dict type15:13
prashanthr_using the ast library15:13
kgriffssriram: ok, cool. so those machines are ready to go? do we also have some load generation box(es)?15:14
prashanthr_is it okay to do this ?15:14
kgriffsprashanthr_: can you paste/gist example?15:14
sriramprashanthr_: eval will help, but its DANGEROUS. on converting a string to dict15:15
kgriffsprashanthr_: i would benchmark to see if msgpack is faster. it is safer15:15
kgriffs(as sriram points out re safety)15:15
prashanthr_I have used msgpack now :)15:15
prashanthr_to store the messages15:15
sriramkgriffs: they are operational, we dont have load generation boxes though15:15
prashanthr_but this is for the queue metadata15:16
prashanthr_will try using msgpack for it as well15:16
kgriffsprashanthr_: yeah, that would be my preference. If someone managed to inject some python code into the db and re read that back and eval'd, that would be A Bad Thing™15:17
kgriffssriram: ok, so next steps15:17
prashanthr_kgriffs: Sure. that's why asked it before the implementation :)15:17
prashanthr_thank you15:17
kgriffssounds good15:18
kgriffsbtw, i have been reviewing your patch, just need to submit my comments so far. I may do it it batches so you get feedback faster15:18
kgriffsprashanthr_: one thing real quick is when using unpackb15:19
kgriffsset encoding='utf-8'15:19
kgriffsand in packb15:19
prashanthr_sure kgriffs. I was doing the latter15:19
prashanthr_now will add the former as well15:19
kgriffsyeah, thing is if you don't set encoding, it won't properly decode six.text_type15:20
*** tonytan4ever has quit IRC15:20
*** tonytan4ever has joined #openstack-marconi15:21
kgriffssix.text_type ---> packb(use_bin_type=True) ---> unpackb() ---> six.binary_type15:21
*** vkmc has joined #openstack-marconi15:21
*** vkmc has quit IRC15:21
*** vkmc has joined #openstack-marconi15:21
kgriffsbut if you set encoding15:22
kgriffsthen you get back out six.text_type15:22
prashanthr_this is really useful ! Did not know that.15:22
kgriffsPITA, but I rest my case15:22
kgriffsprashanthr_: I learned the hard way. :p15:22
kgriffsprashanthr_: one other tiny tip is that if you create the encoder and decoders before hand in the class __init__ or something you can get a little performance boost15:23
kgriffsthat's because unpackb and packb instantiate those classes on the fly under the covers15:24
kgriffsnot a huge perf boost, but thought it was worth mentioning15:24
kgriffssriram: let's make a list of TODO15:24
prashanthr_definitely. so i instantiate the classes and store them as private class variables ?15:25
prashanthr_kgriffs: i guess every call to the library will save one level of lookup :) which could impact performance.15:26
kgriffsonly thing is I'm not sure if those things are thread safe... but we generally recommend people deploy using multiple single threads, and using green thread monkey patches to get async i/o (although at some point we will need to do some extra work to get py3k async)15:28
kgriffsin fact, we should add that to the user guide15:28
kgriffshmm... I wonder if the GIL prevents any badness from happening even when running a server in multithread mode15:29
kgriffsanyway, it's best practice with WSGI apps to do single-threaded procs in most cases, so it isn't like we are asking anything outrageous15:29
sriramasyncio for the win.15:33
*** vkmc_ has joined #openstack-marconi15:43
prashanthr_Have a great day all :)15:44
prashanthr_me going to rest for the day15:44
prashanthr_will upload the latest patch early tomorrow morning15:44
*** cpallares has quit IRC15:45
*** vkmc has quit IRC15:46
srirambe right back15:48
kgriffsprashanthr_: thanks! I will continue reviewing15:50
*** prashanthr_ has quit IRC15:50
sriramkgriffs: looks good15:53
kgriffssriram: let's do 3 load gen boxes15:56
*** whenry has quit IRC16:01
*** flwang__ has joined #openstack-marconi16:29
*** jasondotstar has joined #openstack-marconi16:34
*** flwang__ has quit IRC16:34
*** catherin_ has joined #openstack-marconi16:36
*** catherin_ has quit IRC16:37
*** catherin_ has joined #openstack-marconi16:37
*** catherin_ has quit IRC16:38
*** jasondotstar has quit IRC16:38
*** catherin_ has joined #openstack-marconi16:38
*** jasondotstar has joined #openstack-marconi16:38
*** jasondotstar has quit IRC16:44
*** catherin_ has quit IRC16:46
*** tonytan4ever has quit IRC16:51
*** reed has joined #openstack-marconi16:52
kgriffsmalini1, flaper87|afk: FYI we will need redis >= 2.6 in the gate since that is when Lua scripting landed. looks like trusty has 2.8, so we are good there.16:53
*** balajiiyer has joined #openstack-marconi16:57
*** jasondotstar has joined #openstack-marconi16:57
*** amitgandhi has quit IRC16:58
*** balajiiyer has left #openstack-marconi17:00
kgriffsbbl (lunch)17:01
*** kgriffs is now known as kgriffs|afk17:03
*** flaper87|afk is now known as flaper8717:05
flaper87what about precise?17:06
flaper87malini1: have you had a chance to contact -infra guys about mongodb?17:08
* flaper87 is almost done with sqlalchemy17:08
*** vkmc_ has quit IRC17:12
*** cpallares has joined #openstack-marconi17:14
*** amitgandhi has joined #openstack-marconi17:15
*** amitgand_ has joined #openstack-marconi17:17
*** amitgandhi has quit IRC17:20
malini1flaper87: I did not..Looks like we need to chk if marconi jobs can be allocated trusty17:27
malini1tht will solve our mongo & redis issues17:27
*** whenry has joined #openstack-marconi17:32
malini1So I just chked with infra17:32
malini1It looks like it the witch over to Trusty is fairly easy17:33
*** kgriffs|afk is now known as kgriffs17:33
malini1But we need someone to first confirm tht our unit/functional tests will run ok in Trusty17:33
malini1Can somebody volunteer to do tht?17:33
malini1We need to verify the mongo/redis tests too17:33
malini1I can get you a trusty server, if somebody can do the verification17:36
*** kgriffs is now known as kgriffs|afk17:42
*** whenry has quit IRC17:42
*** vkmc_ has joined #openstack-marconi17:56
*** amitgand_ has quit IRC17:56
*** amitgandhi has joined #openstack-marconi17:57
*** tonytan4ever has joined #openstack-marconi18:03
*** catherin_ has joined #openstack-marconi18:05
*** kgriffs|afk is now known as kgriffs18:08
*** vkmc_ is now known as vkmc18:20
*** vkmc has quit IRC18:20
*** vkmc has joined #openstack-marconi18:20
*** flwang__ has joined #openstack-marconi18:30
kgriffsmalini: I can help if you point me in the right direction.18:31
malini1kgriffs: we need to verify that all our tests can run in trusty18:31
malini1including the mongo & redis(if any)18:31
kgriffsmeaning, devstack?18:32
kgriffsi mean, shall I run it with devstack?18:32
malini1no..just the plain old unit tests18:32
malini1the tests that run with py26, py2718:32
malini1the tests against devstack are tempest tests, which are already on Trusty18:33
kgriffsah, ok18:33
kgriffsrackspace cloud server work OK?18:33
kgriffsI'm on it18:33
malini1thanks kgriffs!18:34
malini1I also need some feedback on
malini1This is to turn on mongo for our devstack/tempest tests18:34
malini1the comment for this one was18:35
malini1   'I'm conceptually good with this, I do think before merging though we need a more concrete error message for the case where the distro provided mongodb is too old, so it's not all cryptic if people run master on precise.'18:35
malini1I don't know how we shud handle this18:35
*** flwang__ has quit IRC18:35
malini1It sounds like we need to eval the mongo DB version & provide the user with an error message18:35
kgriffsthe driver should do that. we should actually have a standard mechanism for all drivers to abort if the version is not supported18:36
kgriffsflaper87: ^^^18:36
malini1so probably we can do this in the devstack script18:36
kgriffsmalini: mmm, we could. how does that compare to doing it in code?18:36
malini1kgriffs: if we can do it easily in the code, tht will be the best option18:37
malini1I just wasn't sure if we could18:37
* flaper87 reads18:37
kgriffssure. we can check and log an error. maybe raise a custom exception type so the bootstrap can abort gracefully18:38
flaper87kgriffs: agreed, I had a similar thought today while working on the sqlalchemy datadriver deprecation18:38
malini1flaper87: can you take care of tht? Tht will help us push the devstack mongodb patch along18:39
*** catherin_ has left #openstack-marconi18:40
*** tonytan4ever has quit IRC18:41
*** tonytan4ever has joined #openstack-marconi18:42
*** amitgandhi has quit IRC18:46
*** amitgandhi has joined #openstack-marconi18:47
*** tonytan4ever has quit IRC18:48
flaper87malini1: yes18:48
flaper87I'll finish the work on merging metadata back into the queue18:48
flaper87and then do that18:48
flaper87pls, there are some patches to review18:48
kgriffsflaper87: have you updated the v1.1 spec yet?18:49
malini1thanks flaper87!! I will follow up on the devstack patch after the mongo version chk18:49
kgriffs(re metadata)18:49
flaper87kgriffs: was going to do that after submitting the patch18:49
kgriffskk, just wanting to make sure that didn't slip through the cracks18:49
flaper87malini1: awesome, thanks.18:49
flaper87kgriffs: make sure you remind me that after the patch hits gerrit18:49
flaper87I'm sure I'll forget18:50
*** tonytan4ever has joined #openstack-marconi18:58
*** saikrishna_ has joined #openstack-marconi19:00
*** saikrishna_ has quit IRC19:00
*** saikrishna has joined #openstack-marconi19:00
*** saikrishna_ has joined #openstack-marconi19:00
*** saikrishna has joined #openstack-marconi19:01
vkmckgriffs, hi Kurt! I chatted with annegentle earlier today, I asked her what we needed for graduate and she gave me this link
vkmckgriffs, I checked the documentation we have and we are missing some things19:09
vkmckgriffs, I didn't ask annegentle to review our docs yet because I thought it was better if we work on adding the changes we know we are missing and then ask her to review all of them19:10
vkmckgriffs, what do you think?19:10
kgriffssounds like a plan19:10
vkmcI'm writting a list of the missing documentation with a structure to organize a all hands docs day19:10
kgriffsok, sounds good. we can rally to get it done.19:11
kgriffsflaper87: ^^^19:11
vkmcbut first I wanted to check with you if you prefer Anne to review the current status of our docs first19:11
kgriffsvkmc: let's do what we know we are missing. and plan one more day to address anything we missed after Anne's review19:11
flaper87kgriffs: awesome19:11
vkmckgriffs, sounds good19:11
flaper87kgriffs: vkmc and myself were talking earlier today about organizing a docs-squashing day next week19:12
flaper87she'll probably send an email later today19:12
* flaper87 keeps talking as vkmc weren't here19:12
* vkmc waves19:12
kgriffsvkmc: can you add some notes to the etherpad?19:13
kgriffsalso, do we need blueprints/or bugs created flaper87?19:13
flaper87vkmc: pls, do send that email today so that folks will read it tomorrow and prepare themselves for next week19:13
flaper87I saw one for docs today, I may be confused though19:14
flaper87I checked earlier today and I think we've all actions covered19:14
flaper87unless I missed some19:14
vkmckgriffs, sure, do you want me to add what we need to add and the workflow to do it?19:14
flaper87vkmc: if we can make it on monday, it would be great. otherwise, whatever day works best for other folks19:14
vkmcflaper87, sure I'll send that email today19:15
vkmcflaper87, I was thinking on Wednesday... we have the meeting on Monday and probably everyone will want to work on they blueprints19:15
flaper87vkmc: ok, sounds good19:15
flaper87I'll be in a conference but I'll do it from there for sure19:16
vkmcplus meetings and other items on theirs agenda19:16
flaper87vkmc: if you break it into bullets, we can get ppl to pick some19:16
vkmcflaper87, oh... are you leaving on Monday?19:16
vkmcflaper87, sure I'm doing that19:16
flaper87vkmc: Tuesday19:16
flaper87I'll attend our meeting19:16
vkmcflaper87, and run away19:17
flaper87naah, just the day after19:17
vkmchaha it's ok19:17
flaper87gtg guys19:18
vkmcflaper87, o/ take care caracolo19:18
*** flaper87 is now known as flaper87|afk19:18
cpallaresvkmc: I would like to help with the docs :)19:21
vkmccpallares, \o/19:21
kgriffsthat would be awesome cpallares if you can help!19:22
vkmccpallares, that would be awesome19:22
vkmcfortunately most of what it needs to be in the developer guide is in the wiki19:22
vkmcand our current developer guide should be transformed in the user guide19:23
kgriffsbtw, don't worry about documenting the api v1.119:23
kgriffswe aren't going to officially release that until k-1 or k-219:23
vkmckgriffs, cool :)19:23
*** oz_akan has quit IRC19:23
vkmcI'll add details in the etherpad
*** saikrishna has quit IRC19:28
*** malini1 has quit IRC19:30
*** jmckind has quit IRC19:38
kgriffsmalini: btw, I am running tox on ubuntu 14.04 now19:51
*** amitgandhi is now known as amitgandhinz19:52
*** amitgandhinz is now known as amitg19:56
*** amitg has quit IRC19:58
*** amitg has joined #openstack-marconi19:58
kgriffsbleh, having trouble with lxml - must be a dep of some module20:07
kgriffstrying to install the right ubuntu packages, stand by...20:07
kgriffsok, working now20:12
kgriffsenabling mongo tests20:12
*** malini2 has joined #openstack-marconi20:25
*** sriram has quit IRC20:29
*** flwang__ has joined #openstack-marconi20:29
*** flwang__ has quit IRC20:31
*** flwang__ has joined #openstack-marconi20:31
*** jmckind has joined #openstack-marconi20:32
*** abettadapur has quit IRC20:35
*** flwang__ has quit IRC20:36
*** openstackgerrit has quit IRC21:01
*** openstackgerrit has joined #openstack-marconi21:02
*** jmckind has quit IRC21:05
*** openstack has joined #openstack-marconi21:08
*** kgriffs is now known as kgriffs|afk21:12
*** amitg has quit IRC21:13
*** tonytan4ever has quit IRC21:21
*** amitg has joined #openstack-marconi21:43
*** mpanetta has quit IRC21:54
*** balajiiyer1 has joined #openstack-marconi21:57
*** whenry has joined #openstack-marconi21:57
*** balajiiyer1 has left #openstack-marconi22:11
*** amitg has quit IRC22:11
*** amitg has joined #openstack-marconi22:12
*** amitg has quit IRC22:16
*** whenry has quit IRC22:25
*** flwang__ has joined #openstack-marconi22:32
*** flwang__ has quit IRC22:37
*** whenry has joined #openstack-marconi22:42
*** ametts has quit IRC22:46
openstackgerritA change was merged to openstack/marconi: Correct the test instructions in README
*** whenry has quit IRC22:56
openstackgerritKurt Griffiths proposed a change to openstack/marconi: Fixes several failing mongo driver tests

Generated by 2.14.0 by Marius Gedminas - find it at!