Wednesday, 2015-12-23

openstackgerritDoug Hellmann proposed openstack/storlets: remove python 2.6 trove classifier
openstackgerritMerged openstack/storlets: Update information about python version in setup.cfg
*** eranrom has joined #openstack-storlets06:45
*** openstackgerrit has quit IRC08:47
*** openstackgerrit has joined #openstack-storlets08:48
*** eranrom has quit IRC09:01
*** eranrom has joined #openstack-storlets09:04
*** openstackgerrit has quit IRC12:17
*** openstackgerrit has joined #openstack-storlets12:18
hrouHey eranrom re: patch 260843 - that's Ok, its in merge conflict because of the other patch that addresses it.13:27
patchbothrou: - remove python 2.6 trove classifier (ABANDONED)13:27
hroui.e. this one: patch 6038713:28
patchbothrou: - Enable hacking check for Apache 2.0 license (MERGED)13:28
*** takashi_ has joined #openstack-storlets13:32
eranromHi. Anybody here?14:00
takashi_eranrom: Hi14:02
eranromHi. Iam I late?14:02
eranromok, I think we should have started an hour earlier, but we have Takashi and Hrou so lets talk :-)14:04
eranromTakashi: any updates, comments?14:04
takashi_eranrom: I have one thing14:05
takashi_I started to add unittests for storlet_middleware, like patch 26033714:06
patchbottakashi_: - Add unittests about validation method in storlet_d...14:06
eranromtakashi_,: yes I have started looking! This is highly appreciated!14:07
takashi_I think we need unitttests before I commit some fixes, because it is useful to make sure tha patch works well14:07
takashi_but I'm afraid that patch is too big, so if neccessory, I'll devide that patch14:08
takashi_into some ones14:08
eranromtakashi_,Don't worry about it. I will go over it and review.14:08
takashi_eranrom: thank you!14:09
eranromI believe that I will have it reviewd by tomorrow EOD14:09
eranromShell we move to talk about Austin Summit?14:09
eranromok, So I wanted to propose that we will submit an NTT-IBM talk on Storlets.14:10
takashi_eranrom: sounds good!14:11
eranromThe general idea is to present your PoC work (if it ok with you) plus work that we have done, show demos do some fuss about storlets.14:11
eranromI am also thinking of trying to arrange a mini desing summit14:11
eranromsay a couple of hours where we describe on more depth the arcitecture and code.14:12
eranromin more depth14:12
takashi_I see14:12
eranromDoes it make sense that I will take a first shot at proposing an abstract and let you review?14:13
takashi_eranrom: Yes for me, but i have to discuss this thing with kota_, first14:13
eranromhrou,: comments?14:14
hroueranrom, that sounds great yea !  Could check if peluse had interest too, but I really like the design summit idea as well to be a bit more hands on / demo14:14
eranromhrou,: Do you think you will be able to take part?14:15
hroueranrom, yep potentially pending ibm travel approval like usual ;)  I can find out though14:15
eranromhrou,: great.14:16
takashi_eranrom: I'll ask kota_ about the idea next Friday, so I think we can tell you some reactions early in the next year.14:16
eranromtakashi_,Great thanks.14:16
eranromAnything else for today?14:17
takashi_do we have some time for discussing about Storlets in the next swift hackathon?14:17
takashi_or should be take another place for that?14:17
eranromtakashi_,: You mean in Bristol?14:17
hrouIt'd be great to talk about that actually !14:17
takashi_hrou: yes14:18
takashi_eranrom: yes14:18
takashi_sorry for mistaking...14:18
eranromhere is what I know: The spec was abandoned by John due to lack on interest at the time. peluse told me that he still thinks we need to push this, but I have not heard from him for a while.14:19
hrouAlways an interesting topic ;) - takashi_ is that what you wanted to discuss (i.e. integrating it as core middleware) ?14:21
takashi_eranrom: I know about the situation. but storlets is still related to swift like swift3. I'm wondering if we can take place at swift hackathon14:21
takashi_I don't think we should integrate storlets to swift now, but think we should keep relations with swift.14:22
eranromtakashi_,: agree14:22
takashi_like keep testing with the latest swift, what we have to do14:22
eranromWe should definately bring up the patch that you wrote14:22
takashi_eranrom: yes14:22
hroutakashi_, yea I think that's a great idea;  Particularly the patch you wrote depending on the status of it (i.e. patch 256201)14:23
patchbothrou: - Fix proxy-server's support for chunked transferrin...14:23
eranromplus I am planning to change s2aio to work with SAIO - that is install SAIO14:23
eranromOnce we have that, we can ask Swift to add a non voting gate that will invoke our tests on Swift commits14:24
takashi_eranrom: great!14:24
hroueranrom, that's a neat idea !14:24
hrouWould have helped catch that EC change that broke it14:25
eranromit will take some time though14:25
takashi_eranrom: I'm looking forward to it14:25
eranromme too :-)14:25
eranromSo you think we should propose a topic on storlets where we briefly update the Swift community on our project?14:26
eranromthat is for Bristol?14:26
takashi_eranrom: that makes sence for me14:27
takashi_honestly speaking, I did not have such kind of idea before talking here,14:28
eranromso this makes it a community idea :-)14:28
hroueranrom, yep!  That's what we did for Austin, kinda of just like a generic update.14:28
takashi_I want to know if we have some time to discuss about Storlets, because I'll decide if i should go to Bristol based on that.14:28
eranromAnyway this is how I interpreted Takashi's idea.14:28
eranromright so lets do that14:28
takashi_eranrom: yeah14:29
eranromoh I see.14:29
eranromin that case we must have that as a topic14:29
eranromand we may want to prepare the 'suggestion' on content together14:29
eranromDo yu want to take a first shot?14:30
takashi_eranrom: I want you to ask john(or other swift members). I think it's better.14:32
eranromAs far as I know the way it works poeple make suggestions and then the suggestions are voted on14:33
eranromSo we can just make a suggestion14:33
eranromdescribing what we want to talk about14:33
hrouyea;  Often they just prepare an etherpad before (probably in a few weeks)14:33
hrouand often just vote there, so we can write whatever we like14:33
pelusehi guys14:34
eranromwhat's up14:34
peluseFYI I'm on vacation for the next two weeks so not super active here :)14:34
peluseSaw the note above though, yeah I don't think things are dead wrt the spec14:34
eranromWe should be on vacation too :-)14:35
peluseFYi I wont be in Bristol :(14:35
takashi_i know that may of you are on vacation now14:35
eranromI am not :-(14:35
peluseyeah, I kinda took a week off the week before last too - heh.  gotta get it all in before EOY14:35
takashi_eranrom: :-(14:35
peluseI think the last storlet thing I posted in here was looking for confirmatin that what I was seeing was expected14:35
peluseseems like months ago :)14:35
eranromno more vacation days for me this year14:35
hroupeluse, hey !  haha yep I recall that14:36
eranromas long as you are back we are happy :-)14:36
peluseI'll fire it up again after the holidays and get back on it.  overall I think the docs will need more work before another core would give things a shot and feel comfortable14:36
eranrompleuse: ok, sounds right. What do you think of a talk / mini design summit on Storlets in Austin?14:37
peluseonce I know my stuff is OK, I need to go through a storlet (now that I think its all installed correctly) and get to the point where I can speak from epxerience to the other guys14:37
peluseyeah, yeah Austin sounds good14:37
hrouyep I think docs are a major area for improvment.  Just for clarity / simplicity14:37
takashi_hrou: absolutely14:37
peluseFYI I've been re-assigned here so Swift is now onlya small part of my job.  I was able to keep 2 aspects alive - the stuff Intel is doing with Rackspace (we have 4 engineers that are part of the cloud for all initiative thingy) and storlets...14:38
pelusefor sure on docs.  I think that's going to be the big thing that will make or break getting the middleware upstream14:38
eranromSounds good.14:39
takashi_great work14:39
peluseBTW the re-assignment is why I didn't even bother asking for $ to go to Bristol.  Austin should be a no brainer though14:39
hroupeluse, Yea that makes sense;  Austin is pretty close for you ;)14:40
eranrompeluse,: Good luck with your new role!14:40
pelusethanks!  Not sure if you've read much on next generation non volatile memory - the byte addressable persistant stuff but that's what I'll be working on14:40
pelusesome light reading here:
eranromsounds interesting (180 degrees from cloud :-)14:41
hroupeluse, oh we read / talk about it all the time (and have some regular calls with intel actually !)  I'll find out who in a sec curious if you know them.14:41
takashi_peluse: sounds exciting for me!14:41
pelusehrou: sure.  this was actually my project before I started on OpenStack stuff so I know the team very well and am excited to work with them again14:42
hroupeluse, wow yea that's an exteremly exciting area to be involved with !14:42
peluseyeah, we may find a way to work this into Swift as well, who knows.  Initially I'll be working with Msft though, specifically to help get that programming model implemented in a way that makes sense for them14:43
hrouYea I was going to say a lot of applications in swift14:44
pelusewe actually have some libraries out there already that help make this stuff more usable:
eranromwho said something about mysql not being optmal for containers14:45
peluseOK, I gotta bolt and take care of way too much pre-xmas stuff I've been putting off... have a great holiday guys!!14:46
eranromThanks, you too. take care14:46
hroupeluse, take care happy holidays!14:47
takashi_peluse: happy holidays14:47
takashi_I think we have to decide the next action for Bristol14:50
takashi_so i'm going to nominate the storlets topic if the proposal becomes available, can I also add your name, eranrom?14:52
takashi_I think everyone knows you're working with Storlets stuff14:52
takashi_if accepted at the hackathon, we can talk there. If not accepted, we can take another place at the week, to talk about storlets topics, including a plan for Austin.14:54
takashi_make sense?14:54
eranromtakashi_,: sure that make sense.14:55
kota_I think we can talk about it as in flight conversation14:55
takashi_kota_: yeah14:56
kota_If needed14:56
eranromkota_,: So you think we should not propose an 'update session' on storlets - that is - specifying we are not talking about pushing this upstream just updating the community14:57
kota_Wait, turn on my laptop.,.14:58
takashi_I suppose you're checking here with your smart phone :-)14:59
kota_takashi_: yup15:00
kota_and reading again via scroll backing15:01
kota_righ now, the conversions for Bristol, right?15:01
takashi_kota_: yes15:01
kota_i saw some talk about Austin summit too tho15:01
takashi_kota_: there are two things. one for Austin and the other for Bristol15:03
hroukota_, we were talking about both ;)15:03
kota_i think that's ok to propose a storlets topic because some one want to catch up current our activities.15:03
kota_s/want/may want/15:03
kota_for Bristol, i was starting to talk here15:03
kota_that is because I had some conversations for swift3 as well as swift :)15:04
kota_in previous Austion hackathon in IBM office.15:04
hrouhehe yep!  I remember that : )15:05
takashi_good news!15:06
kota_i saw the back log paul won't be Bristol? who one will go there?15:07
hrouI think eranrom will be going, not entirely sure my self quite yet15:08
kota_if we have enough members to talk about storlets, i wanna say 'yes we need a meet up there' :)15:08
kota_because it may show a good point of view for other swifters and we may get some feedback for our activities from them.15:09
hrouhaha sure I like that idea i.e. an official swift meetup ; )15:09
hrouYep I completely agree15:09
kota_ok, I will be there and takashi may be able to go there if we have storlet conversation there.15:10
eranromSorry, stepped out for a min. I will be in Bristol15:10
kota_eranrom: no worries and good news to ask my boss to bring takashi to Bristol :)15:10
kota_eranrom: thx!15:10
takashi_kota_: good news for me :-)15:11
kota_takashi: tbh, i'm not sure how my boss think about it :P15:11
hrouhaha !15:11
takashi_kota_: I believe you will help me15:12
kota_will do15:12
kota_and i think15:12
kota_we *may* need to discuss a few swift changes for storlets like chunked transfer on GET.15:13
kota_that's also a reason to bring storlets talk to hackathon.15:13
hrouyep, depending on the state of that by the hackaton15:14
kota_takashi_: enough?15:16
takashi_kota_: yes15:16
kota_fyi,  FEBRUARY 1 IS THE DEADLINE TO SUBMIT A TALK (11:59PM PST) for austion summit breakout sessions.15:17
eranromMay I send you a draft for a mutual talk?15:18
kota_eranrom: takashi?15:19
eranromboth :-)15:19
eranromdepends on you actually15:19
kota_eranrom: great :)15:19
eranromI suggest that we start with the abstract and then we can decide who speaks about what, etc.\15:20
hrouyep that sounds great15:20
takashi_eranrom: sounds good for me15:20
kota_+1 :)15:21
eranromgreat, so expect this in a couple of days :-)15:21
eranromI know you are in holidays so do not fell obligated to answer right away :-)15:21
takashi_eranrom: thank you for your kindness :-)15:22
eranromnot at all. You do not know how exited I am to actually have a storlets community :-)15:23
takashi_happy to hear that! :-)15:25
hrou+100 :)  Thanks all!15:26
kota_ok, I am startting to prepare to head for my bed :)15:28
kota_good night and thanks ;)15:28
eranromgood night. Thanks you all.15:28
takashi_good night15:29
*** takashi_ has quit IRC15:32
*** eranrom has quit IRC17:03
*** eranrom has joined #openstack-storlets19:34
*** eranrom has quit IRC19:36
*** openstack has joined #openstack-storlets22:23
*** openstack has joined #openstack-storlets22:38

Generated by 2.14.0 by Marius Gedminas - find it at!