14:00:15 <nikhil_k> #startmeeting Glance
14:00:18 <openstack> Meeting started Thu Feb 26 14:00:15 2015 UTC and is due to finish in 60 minutes.  The chair is nikhil_k. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:19 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:22 <openstack> The meeting name has been set to 'glance'
14:00:29 <hemanthm> o/
14:00:29 <ativelkov> o/
14:00:31 <kragniz> o/
14:00:39 <nikhil_k> https://etherpad.openstack.org/p/glance-team-meeting-agenda
14:00:48 <stevelle> o/
14:00:50 <mfedosin> o/
14:01:06 <krykowski> o/
14:01:07 <ivasilevskaya> o/
14:01:43 <nikhil_k> #info K-3 FF proposal date is Mar 12 for Glance
14:02:15 <sigmavirus24> o/
14:02:19 <nikhil_k> That means, all the feature related code needs to be proposed & merged by that date
14:02:20 <Olena> o/
14:02:20 <rosmaita> o/
14:02:39 <lakshmiS> o/
14:02:47 <nikhil_k> https://launchpad.net/glance/+milestone/kilo-3
14:03:19 <nikhil_k> Please take a look at that and try to review the linked features + Artifacts on priority
14:03:48 <nikhil_k> mfedosin: ativelkov : Can anyone of you add Artifacts related BPs to k3?
14:04:28 <ativelkov> nikhil_k: sure. Do you want a single artifacts BP or separate BPs for each commit?
14:04:29 <mfedosin> nikhil_k, sure, we're working on it
14:04:48 <nikhil_k> ativelkov: as many specs we've,
14:04:50 <pkoniszewski> o/
14:04:56 <nikhil_k> please update their milestone to be k3
14:05:11 <ativelkov> this means two: one for semver and another one for eveything else
14:05:20 <nikhil_k> yes
14:05:24 <ativelkov> ok
14:05:50 <nikhil_k> (unless discussions from here on call for any other one, I'm trying to be generic)
14:06:09 <nikhil_k> Also, we've had several requests about https://blueprints.launchpad.net/glance/+spec/sighup-conf-reload
14:06:39 <nikhil_k> I just wanted for someone with enough cycles to test that functionally and double check if we did not miss anything
14:07:57 <nikhil_k> #topic Question: grouping artifacts by type version in the list output (as proposed on mini-summit)
14:08:14 <nikhil_k> Who proposed that?
14:08:26 <sigmavirus24> ativelkov: ?
14:08:28 <nikhil_k> ativelkov: looks like you?
14:08:30 <sigmavirus24> (going by color)
14:08:33 <ativelkov> yes
14:08:39 <ativelkov> Just a quick question
14:08:57 <ativelkov> on the mini-summit there was a proposal to group artifacts by the version of their type
14:09:17 <ativelkov> I mean, in the "list artifacts" API output
14:09:38 <ativelkov> in case if the type version is not specified
14:10:08 <ativelkov> This remains an action item on me, so I wanted to clarify if that is a good idea or not
14:10:29 <nikhil_k> I stand +1 on that
14:10:51 <hemanthm> I'm wondering if we necessarily have to do it by default
14:11:01 <ativelkov> Why can't it be done on client side?
14:11:08 <hemanthm> how about we provide a param for grouping
14:11:22 <ativelkov> The problem with it is sorting and getting the marker for pagination iteration
14:12:04 <ativelkov> if this is just a plain list, then the id of the last record is the marker for the next page request
14:12:28 <ativelkov> if the output is grouped then there is no such thing as a "last record"
14:13:15 <nikhil_k> ativelkov: true, that makes things a bit complex
14:13:47 <ativelkov> and the sorting order between items in different group cannot be easily defined if the client wants to compare them
14:14:12 <nikhil_k> ativelkov: ok, how bad would it be to do on glaceclient?
14:14:19 <ativelkov> on the contrary, grouping may be always done on the client-side, as the records contain all the needed fields
14:14:35 <ativelkov> nikhil_k: that's what I wanted to suggest: leave it for client(s)
14:14:37 <sigmavirus24> And sorted() is rather fast once we have all the records
14:14:58 <nikhil_k> ativelkov: ah ok, so supported client and not the consumers. cool
14:15:32 <nikhil_k> sigmavirus24: the worry there is that, people might want to fetch more records out of the DB frquently
14:15:45 <sigmavirus24> nikhil_k: yep
14:16:31 <ativelkov> So, if everybody agrees, then I drop that action item: we'll leave API call to output just list of items, not groups
14:17:18 <nikhil_k> sure, please keep a note on the spec that mentions this drawback
14:17:28 <hemanthm> ativelkov, was there a good usecase for grouping by default?
14:17:50 <ativelkov> hemanthm: I don't know - this was said by somebody at the summit
14:18:16 <ativelkov> I didn't remember the usecase behind that.
14:18:30 <nikhil_k> hemanthm: The general feeling was about usability. Not having to go through all the non relavant artifacts
14:19:00 <nikhil_k> ok, we can discuss the details offline if that's ok
14:19:07 <ativelkov> nikhil_k: if you want to get only the relevant ones, you may always specify type version explicitly
14:19:14 <kragniz> nikhil_k: isn't that where filtering would be more useful?
14:19:33 <ativelkov> e.g. GET /v2/artifacts/someType/v1.2/
14:19:53 <nikhil_k> #info ativelkov to update the Artifacts spec to state that grouping of API results is too complicated due to pagination
14:20:32 <nikhil_k> kragniz: ativelkov : true but if you've multiple ones and are doing something between browsing and searching then this becomes relevant
14:21:18 <nikhil_k> can we move on?
14:22:16 <nikhil_k> I think we can discuss the details offline and/or on ML
14:22:37 <ativelkov> sure
14:22:45 <nikhil_k> #topic Abandoning stale PS from review
14:22:57 <nikhil_k> https://etherpad.openstack.org/p/glance-cleanout-of-inactive-PS
14:23:02 <nikhil_k> http://comments.gmane.org/gmane.comp.cloud.openstack.devel/46352
14:23:38 <nikhil_k> Please reply to that thread if you've something to say
14:24:20 <nikhil_k> Guess jokke_ proposed that and is not around so, may be we'd move on
14:24:31 <kragniz> yeah, move on
14:24:51 <nikhil_k> #topic python-glanceclient release 0.16.0
14:25:04 <nikhil_k> http://osdir.com/ml/openstack-dev/2015-02/msg01858.html
14:25:09 <nikhil_k> http://docs.openstack.org/developer/python-glanceclient/#id1
14:25:50 <nikhil_k> (I think it was jokke_ again who proposed that topic ??)
14:26:13 <sigmavirus24> or pkoniszewski
14:26:22 <nikhil_k> That's all the relevant info, (I think) we need
14:26:29 <sigmavirus24> or krykowski
14:26:31 <sigmavirus24> heh
14:26:39 <nikhil_k> #topic Questionable reviews
14:26:43 <sigmavirus24> Ohai that's me
14:26:46 <nikhil_k> https://review.openstack.org/#/c/159068/
14:27:08 <nikhil_k> sigmavirus24: sure
14:27:11 <sigmavirus24> So https://review.openstack.org/#/c/159068/ is questionable because it presents a serious backwards compatibility issue with the glanceclient
14:27:21 <pennerc> o/
14:27:41 <sigmavirus24> It's also indicative of the fact that we either need to dedicate more time to bug triaging so people don't just pick up non-triaged bugs
14:28:18 <sigmavirus24> The less tangential question here is: Do we even want to require a name for a created image?
14:28:37 <rosmaita> it's weird not to have one, but as long as there's a uuid, who cares?
14:28:40 <sigmavirus24> Commenting on that review/bug would be helpful
14:28:55 <sigmavirus24> rosmaita: right and if the API doesn't require one, why should the client is the bigger question
14:29:15 <nikhil_k> I'd just -2 it as that was the original use case, name is not a required attr so we'd not force it
14:29:51 <rosmaita> sigmavirus24: you are correct, the "official" glanceclient should respect the API, private glance clients can do whatever they like
14:29:54 <sigmavirus24> The bug was never discussed either
14:29:58 <sigmavirus24> rosmaita: right
14:30:18 <sigmavirus24> ativelkov: kragniz other thoughts?
14:30:29 <nikhil_k> sigmavirus24: I do not think that discussions are mandatory on bugs before someone picks it up. Many times the bugs are filed because you're working on something
14:30:34 <kragniz> I agree with rosmaita
14:30:43 <ativelkov> Well, I agree that if the name is not required, then we should not enforce it in glance cli.
14:30:54 <mclaren> seems a bit late for a v1 change of this kind IMHO
14:31:02 <nikhil_k> heh
14:32:03 <rosmaita> looking at the api docs, name is not marked as optional ... but there is no uniqueness requirement, and an empty string is a string, so ...
14:32:12 <ativelkov> I think that this discussion (allowing to create images without a name) happens periodically :)
14:32:12 <sigmavirus24> So the other one, I added before Flavio and nikhil_k chimed in but everyone should still take a look at the bug: https://launchpad.net/bugs/1279832 and review: https://review.openstack.org/#/c/158982/
14:32:14 <openstack> Launchpad bug 1279832 in python-glanceclient "nova image-list can run slowly." [Wishlist,In progress] - Assigned to Shashwat Srivastava (shashwat-srivastava)
14:33:24 <flaper87> o/
14:34:11 <flaper87> why does this keep happening to me?
14:34:16 <sigmavirus24> Anyway, I think we can move on unless people want to discuss those further
14:34:17 <flaper87> sigmavirus24: thanks for the ping
14:34:22 <sigmavirus24> flaper87: you're welcome
14:34:31 <sigmavirus24> I'll ping you more often and regularly in the future =P
14:34:37 <flaper87> sigmavirus24: THANK YOU!
14:34:39 <flaper87> :D
14:34:41 <nikhil_k> hmm, flaper87 good point. may be I need a courtesy ping
14:34:51 <flaper87> nikhil_k: please please please
14:34:54 <flaper87> I'd really appreciate it
14:34:58 <flaper87> I do have a calendar reminder
14:35:02 <flaper87> I just don't have a calendar
14:35:03 <sigmavirus24> we need a meeting bot that pings people in #openstack-glance
14:35:04 <flaper87> :P
14:35:09 <nikhil_k> cool
14:35:11 <sigmavirus24> Just don't create an action item for that =P
14:35:18 <nikhil_k> #topic glance_store release 0.1.11 status
14:35:26 <nikhil_k> http://lists.openstack.org/pipermail/openstack-dev/2015-February/057386.html
14:35:36 <nikhil_k> https://launchpad.net/glance-store/v0/v0.1.11
14:35:40 <nikhil_k> List glance_store 0.1.11 in https://trello.com/b/GFXMXxsP/openstack-glance
14:36:15 * flaper87 clicks and clicks
14:36:20 * flaper87 loves those dogs
14:36:26 <nikhil_k> thanks to flaper87 , that trello board is readable in public (public read access is tested, if you cannot hit it please refresh or try a different tab/browser)
14:36:51 <nikhil_k> that's on on that from me
14:36:58 <nikhil_k> moving on..
14:37:12 <flaper87> :)
14:37:29 <nikhil_k> #topic glance_store release 0.1.12
14:37:37 <nikhil_k> https://launchpad.net/glance-store/+milestone/v0.1.12
14:37:41 <nikhil_k> https://trello.com/b/GFXMXxsP/openstack-glance
14:38:03 <nikhil_k> We usually set a tentative date on the release so that people are aware
14:38:05 <mclaren> kragniz: we should add our socket leak fixes to the 0.16.1 client release?
14:38:15 <flaper87> I need to sync again with cindy to understand the status of the glance_store refactor
14:38:24 <flaper87> but It'd be lovely to have it in 0.1.12
14:38:39 <nikhil_k> If you need some specific ones in a particular release, the dates can be adjusted +2 days
14:38:45 <flaper87> Do we want a glance_store release before the Kilo release ?
14:38:48 <kragniz> mclaren: I was looking at 0.17.0 for those
14:38:57 <flaper87> like 2 weeks after K-3 ?
14:38:59 <nikhil_k> Another reason why the release can be delayed: there aren't enough commits to be released
14:39:07 * flaper87 literally made that up
14:39:21 <mclaren> kragniz: they unbreak nova though, so high priority in my view...
14:39:40 <nikhil_k> One release would be a bit before K3
14:39:58 <nikhil_k> Mar16th or so
14:40:04 <sigmavirus24> mclaren: they're not backwards incompat and they're bug fixes so +1 on 0.16.1 if we can get them reviewed
14:40:08 * flaper87 confused are we talking about the client or glance_store ?
14:40:23 <flaper87> :P
14:40:25 <nikhil_k> flaper87: store
14:40:26 <kragniz> mclaren: yeah, good point
14:40:32 <nikhil_k> flaper87: some are talking about client
14:40:40 * flaper87 rolls eyes :P
14:40:45 <mclaren> sorry, speaking of turn...
14:41:17 <flaper87> re glance_store it sounds good to have 1 or even 2 minor releases before Kilo :)
14:41:42 <flaper87> There are still things that are a must fix before kilo goes out like the chunk/offset handling
14:41:48 <flaper87> zhiyan: has patches for that
14:41:59 <nikhil_k> flaper87: sure  (just trying to not break any gates and keep things smooth at the release week)
14:42:23 <nikhil_k> if we can get those patches reviewed soon, it'd be awesome
14:42:34 <flaper87> I'll make that a priority for me
14:42:47 <nikhil_k> appreciate it
14:42:54 <nikhil_k> #topic Metadef Notifications
14:42:55 <zhiyan> Thanks for you incoming review
14:43:09 <nikhil_k> TravT: wanna go?
14:43:20 <TravT> This review is ready, we think: https://review.openstack.org/#/c/148546/
14:43:36 <TravT> krykowski: updated it last night with outstanding comments
14:43:47 <nikhil_k> nice
14:43:53 <krykowski> yup, its in really good shape ;)
14:44:34 <nikhil_k> krykowski: you might want to check the jenkins failure though
14:44:50 <nikhil_k> people _may_ not review something that has -1 on it already
14:45:06 <TravT> hmm I don't see a jenkins failure
14:45:09 <krykowski> it has +1 from Jenkins
14:45:20 <krykowski> Jenkins is a good guy :)
14:45:35 <nikhil_k> TravT: krykowski : my bad, I confused you for flaper87
14:45:44 <kragniz> a grave mistake
14:45:52 <nikhil_k> sorry about that, we'r good
14:45:57 <TravT> always a good idea to assume flaper87 has jenkins failures?
14:46:03 <TravT> ;)
14:46:05 <nikhil_k> (and I'm asleep)
14:46:26 <nikhil_k> anyways, moving on..
14:46:40 <sigmavirus24> TravT: that was harsh =P
14:46:50 <nikhil_k> #topic glanceclient next few releases
14:47:23 <nikhil_k> TravT: heh, just something that I was looking at and switched the tabs before coming back to IRC window :P
14:47:37 <nikhil_k> https://launchpad.net/python-glanceclient/+milestone/v0.16.1
14:47:48 <nikhil_k> #info This is due next week.
14:48:12 <nikhil_k> https://trello.com/b/GFXMXxsP/openstack-glance
14:48:21 <nikhil_k> card name glanceclient 0.16.1
14:49:01 <nikhil_k> One more is created in LP scheduled to be released a few days before K3
14:49:06 <nikhil_k> https://launchpad.net/python-glanceclient/+milestone/v0.17.0
14:49:36 <nikhil_k> mclaren: Did you have something to mention for that earlier?
14:49:39 <flaper87> TravT: LOL, is that because I always do? :P
14:49:44 <mclaren> nikhil_k: yeah
14:49:51 <sigmavirus24> flaper87: that was the joke ;)
14:49:58 <TravT> flaper87: wondered if you were still listening  ;-)
14:50:05 <mclaren> nikhil_k: can I just add to that launchpad ticket? (What's the process?)
14:50:08 <flaper87> sigmavirus24: no no, I'm serious, I always do :P
14:50:32 <nikhil_k> mclaren: yes, please feel free to add it to LP. We're using the trello board to see if there are any blockers
14:50:34 * flaper87 clicks on the client card
14:50:47 <nikhil_k> LP can get crowded sometimes
14:51:08 <mclaren> https://review.openstack.org/#/c/157516/ and https://review.openstack.org/#/c/156975/
14:51:17 <kragniz> mclaren: do you mean target bugs to a release?
14:51:32 <kragniz> mclaren: you can do it from the launchpad bug page
14:51:37 <mclaren> we found that nvoa and cinder would fall over due to too many open fiel descripters
14:51:38 <flaper87> re client, I need to get back to this: https://review.openstack.org/#/c/140336/
14:51:40 <kragniz> ("target to milestone")
14:51:45 <flaper87> but I don't think it should block the release
14:51:55 <flaper87> so, feel free to move that one to the next release
14:52:03 <flaper87> (if needed)
14:53:10 <nikhil_k> flaper87: please do so, if possible for 0.16.1
14:53:20 <nikhil_k> moving on..
14:53:23 <nikhil_k> #topic Multivalue operator support
14:53:38 <nikhil_k> pkoniszewski: that's you
14:53:43 <mclaren> kragniz: looks like I added them telepathically :-)
14:53:45 <pkoniszewski> thank you
14:53:53 <kragniz> mclaren: that was me :P
14:54:03 <sigmavirus24> mclaren: mind controls kragniz ... good to know ;)
14:54:13 <kragniz> sigmavirus24: it's a problem
14:54:17 * flaper87 wonders who mind controls mclaren
14:54:21 <pkoniszewski> so both patch sets are in a really good shape, actually both had +2 on (had to add two points to spec that zhiyan requested so it dropped)
14:54:43 <pkoniszewski> I'd like to ask for final review there https://review.openstack.org/#/c/147264/ - specification
14:54:43 <sigmavirus24> pkoniszewski: 2 tiny nits from me and then I'm +2 on the spec
14:54:45 <sigmavirus24> (sorry)
14:54:53 <pkoniszewski> and the code - https://review.openstack.org/#/c/148213/
14:55:02 <pkoniszewski> sure, leave them and i'll address it asap
14:55:06 <pkoniszewski> thanks a lot!
14:55:16 * sigmavirus24 is going to look at the code now
14:55:28 <pkoniszewski> thats all for me
14:55:36 <nikhil_k> thanks pkoniszewski, I think based on the momentum we can try to get them in this week
14:55:41 * sigmavirus24 is so glad to see zhiyan doing reviews again
14:55:48 <sigmavirus24> == nikhil_k
14:56:02 <nikhil_k> me too :)
14:56:04 <nikhil_k> #topic Open Discussion
14:56:11 <mfedosin> Folks, I am glad to introduce you our technical writer Olena :)
14:56:11 <zhiyan> For spec ? Or code?
14:56:26 <sigmavirus24> zhiyan: both
14:56:30 <flaper87> *cough* https://review.openstack.org/#/c/159129/ *cough*
14:56:35 <sigmavirus24> Hi Olena !
14:56:38 <flaper87> damnit, gate failure (facepalm)
14:56:38 <Olena> hi Guys!
14:56:46 <zhiyan> Revised spec you have a good eyes
14:56:48 <flaper87> Olena: HEY! :)
14:56:49 <Olena> I am excited to join the Glance team
14:56:51 <sigmavirus24> flaper87: high five jenkins in the face instead ;)
14:57:03 <flaper87> sigmavirus24: good idea, gooooood idea!
14:57:06 <krykowski> Olena: Hi :)
14:57:10 <sigmavirus24> :D
14:57:11 <Olena> =)
14:57:16 <nikhil_k> flaper87: http://giphy.com/gifs/cough-ToLQAiSdwq33G
14:57:22 <rosmaita> Olena: welcome!
14:57:33 <flaper87> nikhil_k: huahuahauhauhau
14:57:43 <zhiyan> Welcome!
14:57:50 <TravT> welcome Olena!
14:57:50 <flaper87> good thing the other 2 are green
14:58:00 <flaper87> (jenkins green, not the other green)
14:58:00 <nikhil_k> Welcome Olena !
14:58:02 <sigmavirus24> I just want to point out that https://review.openstack.org/#/c/140151/ will be unnecessary for any one using requests 2.5.3 or higher but it is still absolutely necessary for anyone using anything less than that
14:58:04 <rosmaita> Olena: maybe you can revise this page: https://wiki.openstack.org/wiki/Glance-where-are-the-docs
14:58:25 <flaper87> sigmavirus24: jeeeez, I wanted to get back to that one last week
14:58:32 <flaper87> I'll review it asap
14:58:38 <sigmavirus24> flaper87: thank you
14:58:47 <nikhil_k> sigmavirus24: yes, thanks for bringing that up.
14:58:50 * sigmavirus24 needs to stop working upstream of openstack so he can complain about himself
14:58:51 <Olena> and I will take a look as well
14:59:41 <nikhil_k> Olena: Thanks for your feedback on the docs that mfedosin shared with us a few days back!
14:59:56 <Olena> oh, my pleasure
15:00:05 <nikhil_k> Olena: please do let know if we can help you get familiar with Glance
15:00:23 <nikhil_k> #openstack-glance is the place we're hiding in :)
15:00:38 <Olena> =) okay
15:00:48 <nikhil_k> ok, thanks all for joining..
15:00:51 <flaper87> o/
15:00:55 <kragniz> nikhil_k: thanks!
15:00:58 <kragniz> Olena: and hi!
15:01:00 <nikhil_k> #endmeeting