08:00:32 <eranrom> #startmeeting storlets
08:00:33 <openstack> Meeting started Tue Jan 31 08:00:32 2017 UTC and is due to finish in 60 minutes.  The chair is eranrom. Information about MeetBot at http://wiki.debian.org/MeetBot.
08:00:34 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
08:00:36 <eranrom> akihito: Hi
08:00:37 <openstack> The meeting name has been set to 'storlets'
08:00:39 <takashi> o/
08:00:43 <eranrom> takashi: Hi.
08:01:06 <eranrom> agenda: https://wiki.openstack.org/wiki/Meetings/Storlets#Agenda:
08:02:20 <eranrom> #topic review-priorities
08:02:27 <eranrom> any comments?
08:03:21 <akihito> nothing
08:03:30 <kota_> can i make sure current stats on review-priorities
08:03:30 <kota_> ?
08:03:39 <eranrom> kota_: sure@
08:03:41 <eranrom> sure!
08:04:04 <kota_> are we.... using that etherpad?
08:04:11 <kota_> https://etherpad.openstack.org/p/storlets-review-priorities
08:04:32 <eranrom> Well, that is the suggestion.
08:04:34 <kota_> sorry, not remember in detail, how it works
08:04:47 <eranrom> It was only put up yesterday - I think
08:04:48 <kota_> probably because of my flu
08:04:56 <kota_> ok
08:05:18 <eranrom> It started off with me asking Takashi to list his own patchers priorities
08:05:31 <eranrom> so he created this pad, and I added my patches there
08:05:55 <eranrom> and now I suggest that we use this, so that each author will prioritize his own patches
08:06:00 <kota_> er, ok
08:06:12 <sagara> hi
08:06:13 <eranrom> so we know as reviewers what to look at first.
08:06:14 <kota_> makse sense
08:06:18 <kota_> sagara: hi
08:06:32 <eranrom> sagara: Hi. agenda is here: https://wiki.openstack.org/wiki/Meetings/Storlets#Agenda:
08:06:47 <eranrom> kota_: So you are more then welcome to add your stuff :-)
08:06:57 <sagara> eranrom: thanks
08:07:28 <kota_> done
08:07:33 <kota_> to add
08:07:35 <eranrom> We can always decide to use something else in the future
08:07:39 <eranrom> kota_: Great! thanks!
08:07:43 <takashi> kota_: :-)
08:08:11 <takashi> as we don't have so much time before Ocata release, we should be careful what we can/should land in this cycle
08:08:24 <kota_> takashi: good point
08:08:41 <kota_> the release should be...
08:08:42 <takashi> and I think that listing can help us, because we can find items with high priorities.
08:08:56 <kota_> the week since 13th Feb?
08:09:06 <eranrom> takashi: right, perhaps this is another subject to discuss. Let me add it to the agenda
08:09:12 <eranrom> kota_: I need to check it...
08:09:33 <takashi> #link https://releases.openstack.org/ocata/schedule.html
08:09:44 <takashi> eranrom, kota_: ^^^^^ fyi
08:09:51 <takashi> but we can talk about this later
08:10:10 <eranrom> takashi: yep! thanks!
08:10:15 <eranrom> next topic?
08:11:02 <kota_> kk
08:11:16 <akihito> yes
08:11:23 <takashi> ok
08:11:30 <eranrom> #topic What should land for Ocata?
08:11:31 <eranrom> :-)
08:12:50 <eranrom> Here is a suggestion:
08:12:55 <eranrom> 1. functional tests
08:13:28 <eranrom> 2. Some of the refactoring work - need to choose.
08:13:35 <eranrom> 3. IPython integration
08:13:43 <eranrom> thoughts?
08:14:25 <takashi> eranrom: Do you plan to have rc in this week?
08:14:30 <takashi> eranrom: or should we?
08:14:43 <takashi> I think that depends on schedule to release rc
08:15:11 <eranrom> takashi: only of you think its important
08:16:05 <takashi> As we are currently a part of official project, so I suppose it is automatically cut off by current release manager
08:16:24 <eranrom> since we do not have any dependencies on our repo (e.g. devstack gateway jobs) I am not sure its important
08:16:42 <eranrom> takashi: In that case not sure it matters, but I may be missing something
08:16:59 <takashi> eranrom: For me it is not so important, but I'm a little concerned a basic rule in OpenStack project
08:17:15 <takashi> as we are now officially a part of it
08:17:20 <eranrom> takashi: ok, I will look into it.
08:18:09 <takashi> For me, it looks better if we can land all func tests work, as we can confirm current behavior by testing
08:18:19 <eranrom> takashi: +1
08:18:35 <kota_> +1, tests is always great
08:18:38 <takashi> as we find some bugs by adding functional tests, like https://review.openstack.org/#/c/426586/
08:19:02 <eranrom> right
08:19:38 <eranrom> Anything beyond the tests?
08:19:53 <takashi> talking about my refactoring work around storlet agents, it is not so critical change because they are just improvoments
08:20:02 <takashi> it is still better, but it is not urgent, I think
08:20:20 <eranrom> takashi: ok.
08:21:03 <takashi> I'll keep working about that, but in some point we need to take a dicision or selection
08:21:07 <takashi> I think
08:21:52 <takashi> and about IPython integration, is there anything remaining? I'm afraid I'm missing some patches.
08:22:17 <eranrom> takashi: There is one patch I did not put yet in the e-pad.
08:22:32 <eranrom> It adds the ability to invoke on get
08:22:41 <eranrom> Let me add it
08:22:45 <takashi> eranrom: this one? https://review.openstack.org/#/c/416089/
08:23:06 <eranrom> yes
08:23:09 <kota_> eranrom: you're doing great pach
08:23:18 <kota_> i'd prefer to review it
08:23:20 <eranrom> kota_: Thanks!
08:23:23 <kota_> but not yet, sorry :/
08:23:26 <eranrom> np
08:23:34 <kota_> just walk through drafting
08:23:37 <eranrom> I take it that you have a backlog
08:23:59 <kota_> thx
08:24:04 <takashi> I'll review it, too
08:24:11 <kota_> AFAIK, it had merge conflicts right?
08:24:20 <kota_> ah, and you already rebased
08:24:27 <kota_> that's what I know right now
08:24:34 <kota_> about the patch
08:25:04 <kota_> and also  https://review.openstack.org/#/c/422511/ (i added as high priority)
08:25:14 <kota_> is preparing ipython work
08:25:28 <kota_> for sort of dry-run
08:26:01 <eranrom> kota_: sure.
08:26:05 <kota_> but imo, it won't happen Ocata timeframe so... in pike?
08:26:18 <eranrom> kota_: yep.
08:27:08 <eranrom> I also added the IPython work but as a low priority :-)
08:27:29 <eranrom> Lets keep high what we want to put in Ocata.
08:27:47 <eranrom> I will move the tests to high
08:27:51 <eranrom> make sense?
08:27:58 <takashi> yes
08:28:04 <kota_> sounds nice
08:28:56 <eranrom> :-)
08:29:12 <eranrom> it was good? why did you delete?
08:29:16 <eranrom> good.
08:29:52 <eranrom> :-) thanks
08:29:54 <takashi> reverting
08:29:57 <takashi> sorry
08:30:25 <eranrom> takashi: Thanks@
08:30:27 <eranrom> Thanks!
08:30:45 <takashi> just added one bug fix: Fixing tenant container id problems
08:30:55 <eranrom> ataraday: yep saw this, thanks!
08:31:19 <takashi> I'll recheck all my patches, and update the list if needed
08:31:26 <takashi> my patches and existing patches
08:31:26 <eranrom> ok great!
08:31:33 <eranrom> takashi: thanks!
08:31:36 <takashi> but currently we can review patches according to it
08:32:04 <eranrom> ok, anythong else on review priorities towards Ocata?
08:32:13 <takashi> nothing else from me
08:33:08 <kota_> current list looks great
08:33:31 <eranrom> ok, that was good thanks!
08:33:44 <eranrom> topic: func tests refactoring thoughts
08:33:50 <eranrom> #topic func tests refactoring thoughts
08:34:38 <eranrom> These are just suggestions...
08:34:50 <eranrom> or thoughts
08:35:52 <takashi> Let me reply to each of your comments.
08:37:18 <takashi> for 1, about temp container usage, I don't know whether we should use temporally container
08:38:16 <takashi> What we would like to care by temporal resource usage is to avoid collisions between multiple test cases
08:38:37 <takashi> especially paralell test execution, which is currently realized by testr
08:38:50 <takashi> (though we are using it in only one thread, currently)
08:39:08 <takashi> s/especially/in especially/
08:39:20 <eranrom> takashi: agree. I guess another question is whether we care about the func test leaving stuff behind...
08:39:50 <takashi> eranrom: Ideally, we should use resources which has unique name, and clean up all things created in test cases
08:40:10 <takashi> s/has unique name/have unique names/
08:40:24 <eranrom> takashi: Agree! It does not have to be done with temp containers.
08:41:38 <takashi> In my func test refactoring, I added some cleaning up about objects created in test cases: https://review.openstack.org/#/c/420419/41/tests/functional/__init__.py@44
08:42:04 <eranrom> takashi: yep. seen that :-)
08:42:37 <takashi> I think it can work if we use temporal container for test cases, but it is not so urgently requied, because collision between containers used in some test cases does not cause so big impart now
08:42:55 <eranrom> takashi: agree.
08:42:57 <takashi> s/impart/impact/g
08:43:26 <takashi> So first step is to use temporal object, and the second step is to clean up all resources created in test execution, I think
08:43:33 <takashi> Does it make sense?
08:43:38 <kota_> !?
08:43:39 <openstack> kota_: Error: "?" is not a valid command.
08:43:55 <eranrom> takashi: absolutely
08:44:09 <takashi> kota_: Do you find any problems?
08:45:02 <kota_> i just was seeing the code a bit
08:45:11 <kota_> but ....
08:45:56 <kota_> i couldn't get why we need *re-implement* methods swift client does
08:46:20 <kota_> it blankets what progress there
08:46:27 <kota_> almost casess
08:47:23 <kota_> and...
08:47:29 <takashi> kota_: but The reason why I'm *re-implementing* them is that I don't get why we need to get self.url and self.token and pass it to client either
08:50:13 <kota_> takashi: why not using swiftclient.Client instance?
08:50:53 <kota_> makking Client instance in setUp, set self.swift_client = Client(brah blah)
08:50:53 <kota_> done
08:52:09 <kota_> and at least, it's not a TestClass but sort of Client class IMHO.
08:53:02 <takashi> kota_: it is still in WIP, so I still need some re-desgin about cleasses
08:53:12 <takashi> and yeah using Client instance makes sense to me. thanks!
08:53:46 <eranrom> We are running out of time and I wanted to spend few minutes on the PTG.
08:53:54 <kota_> ah, it's WIP
08:54:10 <eranrom> perhaps we should discuss more during the PTG itself :-)
08:54:10 <kota_> just was jumping by the url pasted above :P
08:54:21 <kota_> ;-)
08:54:24 <takashi> kota_: yes
08:54:50 <takashi> shortly for 2. yes. it is ideal goal to have same test cases (which inherits same base class which implements common test cases)
08:54:56 <takashi> but still in WIP :-)
08:55:00 <eranrom> takashi: sure!
08:55:14 <eranrom> ok, PTG.
08:55:15 <kota_> takashi: https://review.openstack.org/#/c/422470/
08:55:33 <kota_> takashi: maybe we're going same way
08:55:37 <takashi> eranrom: ok
08:55:42 <kota_> eranrom: sorry, go ahead
08:55:46 <takashi> kota_: I think so
08:56:03 <eranrom> Thanks! dorry for the interruption
08:56:09 <eranrom> #topic PTG
08:56:22 <eranrom> https://etherpad.openstack.org/p/storlets-pike-design-summit
08:56:40 <eranrom> I think it would be good to define goals for Pike
08:57:19 <eranrom> I added two of mine, please add more, and we can discuss durng the PTG
08:57:33 <takashi> eranrom: ok
08:57:40 <eranrom> Otherwise, feel free to add more topics for discussion.
08:58:28 <kota_> eranrom: ok
08:58:41 <eranrom> anything else for the last 2 mins?
08:59:13 <kota_> that's all, great discussion
08:59:21 <eranrom> all thanks very much for joining!
08:59:24 <kota_> thanks eranrom!
08:59:24 <eranrom> talk to you later
08:59:32 <eranrom> #endmeeting