18:00:24 <vgridnev> #startmeeting sahara
18:00:25 <openstack> Meeting started Thu Sep  1 18:00:24 2016 UTC and is due to finish in 60 minutes.  The chair is vgridnev. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:26 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:00:29 <openstack> The meeting name has been set to 'sahara'
18:00:34 <vgridnev> o/
18:00:34 <elmiko> hi
18:00:39 <egafford> o/
18:00:44 <raissa> hi
18:01:13 <vgridnev> let's wait a little
18:01:40 <egafford> tosky's out at the moment.
18:03:26 <vgridnev> #link https://wiki.openstack.org/wiki/Meetings/SaharaAgenda
18:03:32 <NikitaKonovalov> o/
18:03:45 <vgridnev> #topic News / Updates
18:04:47 <NikitaKonovalov> got some progress on boot_from_volume feature
18:05:09 <egafford> Not a ton; I've picked up a ton of internal work again I fear, as is my way. :/ Thanks for pinging me for reviews when things need to move though, vgridnev, it's very appreciated and I very much want to help.
18:05:24 <NikitaKonovalov> both volume and instance create successfully, but attachment fails
18:05:43 <tellesnobrega> o/
18:06:22 <vgridnev> I was preparing N3, making releases, and also did some internal stuff also.
18:06:58 <tellesnobrega> I didn't do much upstream this last week, most internal work. I believe all my patches for N are in, so just keep reviewing for now I guess
18:09:38 <vgridnev> so, how about bugs / docs weeks? Any help at this area is much appreciated.
18:10:41 <tellesnobrega> when is the deadline for bug fixes?
18:10:45 <vgridnev> #topic Open discussion
18:11:22 <egafford> tellesnobrega: RC1 target week is Sept. 12-16, and the release is Oct. 3-7.
18:11:22 <vgridnev> first, just remainder: let's collect topics for discussions at the summit
18:11:26 <vgridnev> #link https://etherpad.openstack.org/p/sahara-ocata-summit
18:12:06 <tellesnobrega> egafford, ideally we want bugs fixed at RC1?
18:12:08 <vgridnev> there are only 1 topic right now, I think that I replace topics from last summit to this etherpad also
18:12:24 <vgridnev> I will replace *
18:12:29 <egafford> vgridnev: Do we have a number of working sessions yet?
18:12:35 <vgridnev> Nope
18:12:43 <egafford> Okey doke.
18:13:01 <vgridnev> I've submitted request about our needs, so now we should just wait
18:13:01 <NikitaKonovalov> egafford: I guess not much since the summint is going to be Tue-Fri
18:13:29 <egafford> NikitaKonovalov: Austin was 4 days too, wasn't it?
18:13:34 <vgridnev> NikitaKonovalov, I hope will have 6wr + half day at least
18:13:58 <NikitaKonovalov> would be nice if so
18:14:03 <egafford> That'd be plenty.
18:16:18 <vgridnev> I think that we should work much better on improving EDP + LCM for clusters in Ocata
18:17:23 * egafford doesn't actually know the acronym LCM, and google is telling me stupid things.
18:17:33 <NikitaKonovalov> yep, I've been on multiple customers meeting recently and almost all of them are asking about cluster management after provisioining
18:17:42 <vgridnev> life cycle management
18:17:44 <NikitaKonovalov> LCM for Life Cycle Management
18:17:56 <egafford> Ah, that's completely sensible, thanks.
18:18:02 <NikitaKonovalov> a separate point in is "Upgrades"
18:18:11 <NikitaKonovalov> everyone wants them
18:18:21 <NikitaKonovalov> and noone does them properly
18:18:34 <egafford> NikitaKonovalov: Yeah, I think we could easily have a whole session on Sahara-side rolling upgrades and a whole separate session on cluster upgrades.
18:18:48 <NikitaKonovalov> egafford: agree
18:19:04 <tellesnobrega> That makes total sense
18:19:56 <egafford> Is there much to talk about in Horizon, in API, or in EDP?
18:20:38 <egafford> API v2 is still hopeful but I feel like we know what we're doing but just don't have effort to put in, so I feel like a full session on that might be less useful.
18:22:00 <elmiko> depending on time, i can try to video phone in for the api v2 discussion
18:22:27 <elmiko> at the least, i can answer any outstanding questions
18:22:50 <egafford> Our GUI is actually really, really good already for an OpenStack UI.
18:23:21 * egafford has summoned elmiko by invoking the mystical phrase "v2 API"
18:23:30 <elmiko> hehe =)
18:23:59 <elmiko> i really do want to see api v2 happen, it just seems like external forces have conspired against me lol
18:24:05 <egafford> If you say it 3 times he appears before you and makes you grow a huge awesome beard.
18:24:10 <elmiko> LOL
18:24:12 <egafford> Yeah, that seems valid.
18:24:37 <elmiko> so, like, i can help get whoever is interested up to speed, and i am willing to do reviews on any of those patches
18:25:27 <vgridnev> egafford, agreed, I'm not sure we have enough resources to work on improving UI, so we just add new features to that (like Boot from Volume, Upgrade)
18:26:10 <egafford> vgridnev: Yeah, I think UI in and of itself is solid and will flow from the featureset, so maybe we don't need a full session there (and can talk upgrades instead or something.)
18:27:17 <egafford> I know our product folks have been starting to talk about rolling upgrades as a big desire in the field, and it's not nearly as hard to do in Sahara as it is elsewhere (since we're all active/active everywhere.)
18:30:07 <egafford> Do we want to start assigning folks to sessions now, or wait on that?
18:30:31 <vgridnev> egafford, so let's wait until we will have number of sessions
18:30:44 * egafford suspects we should get to planning bug days and doc days during this meeting; summit can wait a bit. Yeah, totally agreed.
18:30:55 <vgridnev> ok
18:31:14 <vgridnev> for doc days there is an ether pad https://etherpad.openstack.org/p/sahara-doc-days
18:32:20 <vgridnev> each of us can take doc and start improving that + writing new docs
18:33:51 <vgridnev> I think that most of docs are ok, I also was thinking about some additional screen casts on how everything works
18:34:10 <egafford> Cool. We can just note assignee and status on the pad. For statuses - {assigned, in progress, blocked (with description), under review, done}?
18:35:01 <egafford> vgridnev: Yeah, we're not in terrible shape. Screenshots sound great, though they're always a little subject to change.
18:35:07 <tellesnobrega> vgridnev, I love that idea. One of the hardest thing to know is what goes where, screencasts are very handy for that
18:35:19 <egafford> Still, I think the UI should be more stable now.
18:35:24 <tellesnobrega> screenshots and casts
18:35:48 <egafford> (The consolidation of clusters and jobs panels was really good, and makes things really pretty intuitive.)
18:36:34 <tellesnobrega> Sure, but I still believe that some helpful graphic guides are very useful
18:36:50 <egafford> vgridnev: Do you want to run point on the bug days and I'll run point on the docs days? (Or would you like someone else to step up for bug days, since you're a busy man? :) )
18:36:50 <vgridnev> for bugs: the current status of RC1 is following: https://launchpad.net/sahara/+milestone/newton-rc1
18:37:38 <egafford> Nice!
18:38:19 <vgridnev> egafford, since you are native speaker, I think that it's much better if you will be handling doc days
18:38:57 <vgridnev> about bugs, additionally there few bugs that currently out of milestones
18:39:03 <egafford> vgridnev: Yeah, that was my thought.
18:39:03 <egafford> re: bugs: https://bugs.launchpad.net/sahara/+bugs?orderby=-importance&start=0 I'm seeing a few bugs without milestones that are critical or high prio.
18:39:56 <egafford> We may want to target these so we can use the handy newton-rc1 link to keep things really nice and uncluttered during bug days.
18:40:43 <egafford> https://bugs.launchpad.net/sahara/+bug/1532826 in particular is listed as critical.
18:40:43 <openstack> Launchpad bug 1532826 in Sahara "sahara-all leaks memory" [Critical,Triaged]
18:41:34 <egafford> 6 high prio bugs too.
18:42:12 <vgridnev> about this one, it's a quite interesting, and it require a lot research
18:42:31 <vgridnev> I was talking about sahara-all leaks memory
18:42:54 <NikitaKonovalov> it reproduces only under very large scale circumstances
18:43:05 <NikitaKonovalov> like 200+ vms in a cluster
18:43:27 <egafford> vgridnev: Yup, it looks it.
18:43:48 <egafford> NikitaKonovalov: Hm; interesting.
18:44:31 <vgridnev> actually, it's seems like sahara leaks not just SSH keys, it leaks objects like clusters data on so on
18:44:38 <NikitaKonovalov> I've tried investigating some time ago. I had a few suspects lie sqlalchemy objects or our own conductor objects leaking. But none of that proved to be true.
18:47:13 <egafford> Hm; so that sounds like it is critical, but also not necessarily appropriate to target to RC1.
18:47:26 <vgridnev> there is also interesting issue like https://bugs.launchpad.net/sahara/+bug/1605134
18:47:26 <openstack> Launchpad bug 1605134 in Sahara "[CDH]Fail to update cloudera password" [High,Confirmed]
18:47:27 <egafford> Any of the High prio bugs look like RC1 candidates?
18:48:57 <vgridnev> so, it's quite important when namespaces are used, because we can't just use requests or something like that to open HTTP client
18:49:51 <vgridnev> about embedded jars, I think that we should downgrade that to Medium and then fix that in ocata
18:50:18 <vgridnev> https://bugs.launchpad.net/sahara/+bug/1609870
18:50:18 <openstack> Launchpad bug 1609870 in Sahara "Embedded hadoop-swift connector in mapr plugin code" [Medium,Confirmed]
18:50:44 <vgridnev> related issue to CDH: https://bugs.launchpad.net/sahara/+bug/1600666
18:50:44 <openstack> Launchpad bug 1600666 in Sahara "cdh plugin uses manager_ip to check cloudra-manger service" [Medium,Confirmed]
18:51:34 <egafford> I know tosky's really excited to get that fixed, but I can see the decision.
18:52:20 <egafford> That CDH bug is kind of nasty really.
18:53:07 <vgridnev> egafford, so, if he really want to get that fixed, we can consider that to RC1. I'll contact him about that
18:53:36 <egafford> Yeah, let's keep the embedded .jars at RC1 for now; we can always punt them out to O later.
18:54:26 <vgridnev> I think that we can close 1600666 as duplicate of 1605134
18:54:46 <vgridnev> these bugs are almost around same issue
18:56:03 <tellesnobrega> vgridnev, yeah, seems like so. Are you sure that if I fix one the other will be fixed?
18:56:09 <egafford> vgridnev: Yeah, it sounds like one reporter just dug a little deeper to find out *why* it was failing.
18:56:29 <tellesnobrega> if we are sure on that we can close
18:57:14 <tellesnobrega> quick time check - we got 2 minutes
18:57:39 <vgridnev> tellesnobrega, yes, we can use some simple action on manager to verify it's active
18:57:57 <vgridnev> I don't know, list of clusters, for example
18:58:04 <tellesnobrega> vgridnev, cool.
18:58:14 <tellesnobrega> we can think about the fix later :)
18:59:16 <vgridnev> Ok, seems like we have 1 min, I think that we can continue at openstack-sahara
18:59:21 <vgridnev> thanks all
18:59:26 <egafford> o/
18:59:28 <vgridnev> #endmeeting