14:00:07 <tellesnobrega> #startmeeting sahara
14:00:07 <openstack> Meeting started Thu Jan 18 14:00:07 2018 UTC and is due to finish in 60 minutes.  The chair is tellesnobrega. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:09 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:12 <openstack> The meeting name has been set to 'sahara'
14:00:16 <tosky> o/
14:00:22 <jeremyfreudberg> o/
14:00:43 <tellesnobrega> waiting for shuyingya to show hands
14:01:06 <jeremyfreudberg> i'll need to leave at 45-50ish minutes past
14:01:21 <tellesnobrega> ok, we might not have too long meeting today
14:01:29 <tellesnobrega> #topic News/Updates
14:02:35 <tosky> cleaning few reviews, sent the review which enables the build and upload of oozie on tarballs (please review!), some various testing
14:03:05 <shuyingy_> o/
14:03:33 <tellesnobrega> I'm working on file transfer bug that we talked about in Denver, during the work I found out that we have a swift - keystone v3 problem that I found a quick solution, I will send a patch soon on that
14:04:39 <jeremyfreudberg> not much new to say, just trying to wrap up the client stuff (1 week till client freeze) and also finishing up force delete
14:05:17 <jeremyfreudberg> oh right, and s3_hadoop fixes
14:05:28 <shuyingy_> working on image build bugs and xenial image python issues as jeremyfreudberg suggested.   also fix downstream tempest tests etc, it also affects upstream
14:06:43 <tellesnobrega> sounds we all have our hands full, we have 1 week until Q3
14:07:08 <tosky> and then bugfixes all way
14:07:20 <tellesnobrega> yes
14:07:38 <tellesnobrega> anymore news?
14:08:03 <jeremyfreudberg> purposely trying to not turn my news into a big question/discussion
14:08:16 <jeremyfreudberg> please proceed
14:08:25 <tosky> oh, it will be the next point I suspect
14:08:30 <tosky> :)
14:08:33 <tellesnobrega> cool (rushing so we can fit all we need)
14:08:46 <tellesnobrega> #topic APIv2 (a.k.a jeremy's show and tell)
14:08:56 <tellesnobrega> tosky, you have the floor
14:09:01 <tosky> I have a question in fact
14:09:13 <jeremyfreudberg> go ahead
14:10:05 <tosky> it's a follow up to the discussion that we had yesterday
14:10:58 <tosky> the conclusion was that the form of the endpoints does not matter, and that we should update the devstack code with the unversioned endpoint after the saharaclient patch lands
14:11:42 <tosky> and only the unversioned endpoints, if I understand it correctly
14:11:42 <jeremyfreudberg> the unversioned endpoint should come when a v2-compatible saharaclient is on pypi
14:11:53 <tosky> yes, but what about the old client?
14:12:07 <tosky> it is true that the new saharaclient will be able to deal with the new endpoint
14:12:27 <tosky> but shouldn't we make sure that old clients still continue to work with v1 if the old endpoints are exposed?
14:12:33 <jeremyfreudberg> if you remember, monty's email suggested old deployments keep versioned url in catalog until a long time from now
14:12:37 <jeremyfreudberg> new deployments have the unversioned
14:12:51 <jeremyfreudberg> when i talk about devstack, i treat it as a fresh (not long lived) deployment
14:13:47 <tosky> that's an interesting question, but given that sahara/queens will depend on the new saharaclient, it should not be a problem
14:14:15 <tosky> but on the other side we should check grenade, so that the new endpoints are added too
14:14:35 <tosky> I don't know the internals of grenade too much, and our jobs are still failing now, but I hope to have them fixed for the PTG
14:14:52 <jeremyfreudberg> (btw, i'm assuming that vendors will provide the latest saharaclient with queens, if we don't get our minimum requirements bump in)
14:15:20 <jeremyfreudberg> yes - keep an eye on grenade for sure
14:15:32 <tosky> yes, that should happen; I was just concerned about the backward compatibility
14:15:45 <tosky> but it seems it should work then - thanks
14:16:21 <jeremyfreudberg> yep
14:16:42 <tellesnobrega> thanks for clarification jeremyfreudberg
14:17:09 <tellesnobrega> any more on this topic?
14:17:19 <tosky> should we ping mordred? :)
14:17:27 <tellesnobrega> probably
14:17:30 <jeremyfreudberg> we should
14:17:35 <tellesnobrega> already did I guess
14:18:03 <jeremyfreudberg> i mean, the client will work as it is but i'm sure mordred would prefer the endpoint code to be more proper there
14:18:29 <tosky> and to easy his search in the log, we are talking about https://review.openstack.org/#/c/533066/
14:18:47 <tellesnobrega> yes, thanks tosky was getting that
14:19:10 <jeremyfreudberg> indeed, it's a bit early in most of the USA, he's probably not around yet
14:19:33 <tellesnobrega> since mordred is not around right now, we can move on and when available we can talk @ #openstack-sahara
14:19:45 <jeremyfreudberg> yep
14:20:19 <tellesnobrega> #topic Dublin PTG
14:20:45 <tellesnobrega> I don't know if you all have seen, but we were allocated Thursday and Friday in Dublin
14:21:45 <tellesnobrega> sadly I discovered that my flight leaves Dublin at 17:40h, so I probably won't be able to be around Friday afternoon, but I believe it should be enough time
14:22:07 <tosky> you can leave shortly after lunch
14:22:17 <tosky> around 14:00 should be enough
14:22:34 <tellesnobrega> yes
14:22:36 <tellesnobrega> that is the plan
14:22:41 <jeremyfreudberg> yes, i think we can squeeze enough into that day
14:23:04 <tellesnobrega> whatever is left I'm more than sure that tosky and jeremyfreudberg and work it out
14:23:24 <tellesnobrega> also, we have the etherpad up, lets start feeding it
14:24:26 <tellesnobrega> that closes PTG, any questions there?
14:25:00 * tosky has no questions for now
14:25:11 <jeremyfreudberg> we can move on
14:25:23 <tellesnobrega> #topic Open Discussion
14:25:49 <tellesnobrega> I guess that is the last topic we have
14:26:25 <tosky> the queue of incoming reviews is better, and all the tiny doc patches have been reviewed and were merged or waiting for some action from the submitter
14:26:43 <tosky> well, some of them are waiting a +2
14:26:51 <jeremyfreudberg> yes, a lot of small patches recently
14:27:02 <jeremyfreudberg> i guess people want to get their name in before freeze
14:27:22 <tosky> we are running out of http and spelling errors :)
14:27:30 <tellesnobrega> getting very close
14:28:15 <jeremyfreudberg> the other thing to mention, about s3_hadoop
14:28:28 <tosky> after the oozie generation (thanks! let's see if it does what it should), now I'm personally waiting for the +1->+2 by jeremyfreudberg on the MapR review
14:28:36 <tosky> oh, right, s3_hadoop
14:29:00 <jeremyfreudberg> i think i need to fix it in two patches, right? one which is the minimal fix that we can backport to pike
14:29:14 <jeremyfreudberg> and finally release 7.0.1 (which is desperately needed downstream)
14:29:22 <jeremyfreudberg> and then a bigger fix for master only
14:29:30 <tellesnobrega> sounds about right
14:29:36 <tosky> yes, the patch to fix vanilla, we have a bug for that
14:29:50 <tosky> that's more urgent; the rest can be postponed between queens-3 and the final release
14:29:54 <tosky> in the bugfix phase
14:30:06 <jeremyfreudberg> yes, exactly
14:30:48 <jeremyfreudberg> my point is: we can backport something easy to pike that might not totally support s3 functionality but just gets clusters to work
14:31:22 <tellesnobrega> yes
14:31:47 <tosky> yes
14:31:50 <jeremyfreudberg> btw, shuyingy_ what is happening with hbase element for vanilla image?
14:33:15 <shuyingy_> I don't have hands to take care of vanilla upgrade and habse integration.
14:34:04 <shuyingy_> vanilla upgrade is very close to finish. I will update them asap
14:34:22 <tellesnobrega> shuyingy_, I would say focus on vanilla upgrade
14:34:28 <tellesnobrega> we need that asap
14:34:31 <jeremyfreudberg> yes, i would say vanilla upgrade is priority
14:34:50 <jeremyfreudberg> we could get FFE for hbase maybe
14:34:57 <tellesnobrega> probably
14:35:07 <tellesnobrega> depends on its state and probability to break stuff
14:35:12 <shuyingy_> vanilla has last step to push is that oozie webUI need two jar packages conflict with tomcat.
14:35:40 <jeremyfreudberg> java dependency hell is not fun
14:35:51 <tellesnobrega> never
14:35:59 <shuyingy_> I will remove them when starting cluster.
14:37:01 <tellesnobrega> is that the best solution?
14:37:06 <shuyingy_> FYI: The method getJspApplicationContext(ServletContext) is undefined for the type JspFactory.
14:37:30 <shuyingy_> I have tried to remove it from  SIE.
14:38:47 <tellesnobrega> didn't work I suppose
14:38:49 <shuyingy_> but found it doesn't have these files. these files are decompressed when executing "oozied.sh start"
14:38:59 <jeremyfreudberg> shuyingy_, makes sense
14:38:59 <shuyingy_> yep
14:39:34 <tellesnobrega> ok, that makes sense then
14:39:38 <shuyingy_> so we have to remove /opt/oozie/oozie-server/webapps/oozie/WEB-INF/lib/servlet-api-2.5.jar and jsp-api-2.0.jar from sahara side
14:40:01 <tosky> s/dependency hell /
14:40:19 <shuyingy_> crazy :(
14:40:46 <tellesnobrega> let us know if it works
14:40:55 <shuyingy_> the dawn is coming
14:41:08 <shuyingy_> it works now.
14:41:14 <tellesnobrega> awesome lol
14:41:18 <jeremyfreudberg> \o/
14:41:39 <shuyingy_> that's all
14:42:14 <jeremyfreudberg> another quick topic: is anyone proposing for the next summit?
14:42:17 <jeremyfreudberg> in may
14:42:20 * tosky is not
14:42:36 <tellesnobrega> not right now, do you have anything in mind?
14:43:24 <jeremyfreudberg> not sure if i will propose anything, but i do have a lot to say about part-time contributions and attending university concurrently, stuff like that
14:43:33 <tellesnobrega> cool
14:43:41 <tellesnobrega> it would be interesting
14:43:46 <tellesnobrega> I've done my share of that
14:44:11 <jeremyfreudberg> good way to advertise myself as well
14:44:12 <shuyingy_> +1 would be interesting
14:44:19 <tellesnobrega> of course
14:44:58 <tellesnobrega> I'm thinking on topics, but nothing came to mind yet
14:46:29 <tellesnobrega> anything else?
14:46:52 <tellesnobrega> I think jeremyfreudberg is almost leaving, if we don't have anything more to discuss I will close the meeting
14:47:05 <jeremyfreudberg> yes - i am almost leaving
14:47:18 <tosky> and we are not far from the end of the hour anyway
14:47:41 <tellesnobrega> yes
14:48:02 <tellesnobrega> lets keep up the work, we have 1 week to finish all features.
14:48:18 <tellesnobrega> see you all around
14:48:25 <tosky> o/
14:48:31 <jeremyfreudberg> bye
14:49:12 <shuyingy_> bye
14:49:41 <tellesnobrega> #endmeeting