16:00:10 <dims> #startmeeting oslo
16:00:10 <dims> courtesy ping for GheRivero, amotoki, amrith, bknudson, bnemec, dansmith, dhellmann, dougwig, e0ne, flaper87, garyk, harlowja, haypo,
16:00:10 <dims> courtesy ping for ihrachyshka, jd__, jecarey, johnsom, jungleboyj, kgiusti, kragniz, lifeless, lintan, ozamiatin, redrobot, rpodolyaka,
16:00:11 <openstack> Meeting started Mon Oct 19 16:00:10 2015 UTC and is due to finish in 60 minutes.  The chair is dims. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:12 <dims> courtesy ping for sergmelikyan, sreshetnyak, sileht, sreshetnyak, stevemar, therve, thinrichs, toabctl, viktors, zhiyan, zzzeek
16:00:13 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:15 <rpodolyaka1> o/
16:00:16 <openstack> The meeting name has been set to 'oslo'
16:00:20 <ozamiatin> o/
16:00:22 <bknudson> hi
16:00:28 <kgiusti> o/
16:00:28 <johnsom> o/
16:00:31 <kzaitsev_mb> o/
16:00:56 <abhishekk> o/
16:01:11 <dims> just a few of us :) hopefully quick meeting then!
16:01:13 <jecarey> o/
16:01:13 <dims> #topic Red flags for/from liaisons
16:01:30 <bknudson> nothing for keystone
16:01:41 <dims> so, last week oslo releases stepped on liberty branches, everyone recovered from it?
16:01:43 <johnsom> We had a missing dependency that the removal of testresource from oslo.db uncovered, but otherwise nothing
16:01:49 <dims> right thanks johnsom
16:02:07 <bknudson> keystone isn't entirely recovered but things are working now.
16:02:48 <dims> bknudson: what can i help with?
16:03:19 <amrith> ./
16:03:30 <bknudson> I think I just need to recheck https://review.openstack.org/#/c/235923/ now that openstack/requirements is working, then https://review.openstack.org/#/c/235923/ needs to get merged  -- dims
16:03:40 <dims> hi amrith, trove recovered from last weeks releases?
16:03:47 <amrith> dims, not entirely
16:03:50 <amrith> hello all
16:03:53 <bknudson> then we'll nee to merge the requirements update and release ksm
16:04:02 <jungleboyj> Hello all.
16:04:02 <amrith> for context, the trove fire was oslo.utils 2.6.0 causing tox tests to hang
16:04:06 <amrith> not enturely sure what the cause was
16:04:11 <amrith> still looking into it
16:04:13 <dims> bknudson: k, ping me when you get stuck
16:04:21 <amrith> there was a change that was proposed
16:04:31 <amrith> https://review.openstack.org/#/c/235149/
16:04:41 <amrith> that hasn't had the expected results (in the most recent incarnation)
16:05:01 <amrith> other than that, we're back in action
16:05:21 <amrith> I do have an issue (not a red flag) for later in the meeting.
16:05:30 <amrith> it relates to oslo.messaging.
16:05:45 <dims> amrith: local test work ok?
16:05:57 <amrith> local tests don't work either
16:06:03 <amrith> i.e. not just gate
16:06:09 <amrith> local machine also hangs with that oslo.utils issue
16:06:12 <amrith> that's how I found it
16:06:21 <amrith> was that the question?
16:06:40 <dims> amrith: with the proposed fix from josh, are you able to pass local tests?
16:07:05 * amrith scrolls back in aging memory
16:07:14 <amrith> yes, when this problem was reported
16:07:21 <amrith> there was a patch proposed to oslo.utils
16:07:23 <amrith> which I tested
16:07:25 <amrith> and it worked.
16:07:26 <dims> amrith: ack, that's good news
16:07:37 <amrith> so when 2.6.1 comes out (or whatever the numebr is) we'll be good
16:07:44 <dims> cool
16:07:45 <amrith> and we can revert our change pinning 2.5.x
16:07:49 <amrith> what's the ETA on that?
16:07:56 <dims> which takes us to next topic
16:07:59 <dims> #topic Releases for Mitaka
16:08:00 <dims> #link https://review.openstack.org/#/c/236770/
16:08:00 <dims> Releases today after verifying tests this morning https://travis-ci.org/dims/
16:08:28 <dims> amrith: waiting for dhellmann's ack on the review https://review.openstack.org/#/c/236770/
16:08:37 <dims> will cut releases after that
16:09:03 <dims> here's the list of all the changes that will be going in - http://logs.openstack.org/70/236770/1/check/gate-releases-tox-list-changes/2a9827b/console.html
16:09:28 <dims> grep for "Details Contents" and you will see a list for each project
16:09:33 <amrith> ok, thanks. I will relay to trove team
16:09:56 <dims> k switching topic
16:09:59 <dims> #topic openstack spec reviews - oslo.apiclient pika rabbitmq driver
16:09:59 <dims> #link https://review.openstack.org/#/c/228992/ (pika)
16:09:59 <dims> #link https://review.openstack.org/#/c/235200/ (apiclient)
16:10:37 <dims> short summary - apiclient going in as a new library, pika driver will be on a branch per what we talked a week or two ago
16:10:52 <dims> any objections?
16:11:11 <abhishekk> hi, I am working on apiclient liabrary
16:11:20 <bknudson> I really think that we should get jamielennox to look at the apiclient library
16:11:46 <abhishekk> please let me know if you have any suggestions
16:12:01 <dims> bknudson: can you please pass on the message?
16:12:11 <abhishekk> bkundson: I will try to ping jamie on this
16:12:17 <dims> abhishekk: thanks
16:12:22 <bknudson> http://git.openstack.org/cgit/openstack/oslo-incubator/tree/openstack/common/apiclient/auth.py#n20 -- this is deprecated, so why put it in a library?
16:12:54 <dims> bknudson: the idea is to start from oslo-inccubator code, then stabilize the API then make releases
16:13:10 <abhishekk> dims: ++
16:13:22 <dims> so it's quite possible that the code in question will be ripped out
16:14:30 <dims> ok next topic
16:14:34 <dims> #topic need drivers for each session
16:14:34 <dims> #link http://mitakadesignsummit.sched.org/overview/type/oslo
16:15:15 <dims> please let me know which session you would like to chair :) or i'll assign one to you :)
16:15:34 <dims> seriously, need help please let me know
16:15:59 <dims> bknudson: can you make the ossg/log session?
16:16:48 <bknudson> dims: that one I can make as of now.
16:16:55 <dims> bknudson: cool thanks
16:16:59 <ozamiatin> dims: I forgot which working session relates to zmq?
16:17:01 <dims> let me know if that changes
16:17:18 <dims> ozamiatin: the 5:20 pm
16:17:36 <dims> ozamiatin: let me pin you to that session :)
16:17:40 <ozamiatin> dims: got it, so it would be pika + zmq?
16:17:57 <dims> ozamiatin: and kafka if anyone from there is around
16:18:25 <dims> ok last item on agenda
16:18:28 <dims> #topic Open discussion
16:18:28 <dims> Any stuck reviews? or specs?
16:18:40 <amrith> I have one
16:18:44 <amrith> or three ...
16:18:50 <dims> amrith: shoot
16:19:07 <amrith> item #1: We (Trove) keep getting requests from people who would like a different messaging transport. Trove uses oslo.messaging and people have suggested we should use Zaqar. I believe that if that is the case, we need a Zaqar driver for oslo.messaging so trove doesn't need to implement code for different RPC systems. What are the plans for something like this?
16:19:08 <johnsom> Slightly off topic, but 4.0.1 of coverage is out and fixes the regression for os.path.exists
16:19:52 <kzaitsev_mb> ok, So I've had a strange idea this weekend, that I'd like to share with you guys
16:19:56 <amrith> Also related to this, as Trove uses oslo.messaging, it is not nice that we have to then go and configure users and ACL's directly with the transport. So I'd like to propose that oslo.messaging provide basic user and ACL API's. Create user, set exchange access ...
16:20:06 <amrith> that's my item #1
16:20:07 <dims> johnsom: nice, thanks. will check if upper-constraints is updated to match
16:20:20 <amrith> or 1, 2 and 3 depending on how you count
16:20:36 <dims> amrith: we have a session for you :) http://mitakadesignsummit.sched.org/event/e415aa81e64ec400204e6554b4cbc249
16:21:19 <dims> amrith: i'll start doing the etherpads and we an throw these ideas in there
16:21:37 <dims> kzaitsev_mb: what's up?
16:22:02 <kzaitsev_mb> So my idea came from realization, that no one in OpenStack uses sentry, I mean https://www.getsentry.com/welcome/
16:22:29 <dims> johnsom: http://git.openstack.org/cgit/openstack/requirements/tree/upper-constraints.txt#n85 yay
16:22:59 <kzaitsev_mb> it is a nice python tool for error logging and arror log aggregation, that's capable of showing nice stacktraces and local variables and stuff
16:23:43 <kzaitsev_mb> So I imagined that one could install devstack and, say, `enable_plugin sentry`
16:23:53 <dims> kzaitsev_mb: central reporting from a customer environment?
16:24:23 <kzaitsev_mb> and have a working sentry in the env, and any 500 error (or probably 40x too) would show up there
16:24:40 <mdbooth> dims: Assume I don't have to ask for additional eyes on my oslo.messaging locking patches? You and harlowja are on it, right?
16:24:51 <amrith> ok, thanks dims. I'll let the others finish their discussion. I have one last question.
16:24:54 <dims> kzaitsev_mb: kind of like logstash usage in our CI
16:25:04 <dims> mdbooth: ack
16:25:08 <dims> mdbooth: yes
16:25:17 <mdbooth> Cool, thanks.
16:25:28 <kzaitsev_mb> dims: haven't looked into that one yet.
16:25:46 <abhishekk> dims: I have one 0/
16:25:51 <dims> kzaitsev_mb: pretty easy to do a plugin for devstack, what would you like our help with?
16:26:08 <dims> abhishekk: shoot
16:26:15 <abhishekk> graduate.sh script fails to delete openstack folder as it contains _i18.py
16:26:25 <abhishekk> * _i18n.py
16:26:40 <abhishekk> is it a bug or i missing something??
16:26:46 <dims> kzaitsev_mb: see http://logstash.openstack.org/
16:26:50 <kzaitsev_mb> dims: yep, I was thinking more of a — should I try adding a oslo-based interface for sentry. That would probably allow other projects to easier adopt it
16:27:40 <dims> kzaitsev_mb: commercial vendor?
16:28:11 <dims> kzaitsev_mb: don't know what the interest level would be. worth starting an email thread to see who else would be interested
16:28:16 <dims> kzaitsev_mb: ok?
16:28:28 <rpodolyaka1> kzaitsev_mb: most operators would probably want to use their own instance of logstash/elasticsearch/kibana rather than relying on  sentry
16:28:33 <dims> abhishekk: possibly, we haven't run that thing in ages
16:28:40 <kzaitsev_mb> dims: nothing commercial there )
16:29:03 <amrith> quetsion #4 ... Can I configure oslo.log so it redirects all messages to syslog on a different host?
16:29:08 <abhishekk> dims: ok, I will report it as a bug
16:29:13 <dims> kzaitsev_mb: looking at https://www.getsentry.com/pricing/
16:29:29 <kzaitsev_mb> dims: rpodolyaka1: sentry can be self hosted
16:29:57 <dims> amrith: i haven't done it! dhellmann may know
16:30:00 <kzaitsev_mb> it's an open source python tool. It's just that they sell it as a service
16:30:13 <rpodolyaka1> kzaitsev_mb: I must be missing something, looks like they offer a SaaS there
16:30:23 <haypo> quick poll for documenting changes in oslo (see also my thread on openstack-dev ML), please vote:
16:30:27 <haypo> +2 to document them in reviews and make them mandatory for new changes
16:30:32 <haypo> +1 to document them in reviews (but keep them optional, we may document changes later)
16:30:37 <haypo> 0 if you don't care
16:30:38 <kzaitsev_mb> rpodolyaka1: https://github.com/getsentry/sentry here is the source code
16:30:38 <haypo> -1 if you don't want to document changes in reviews
16:30:38 <dims> kzaitsev_mb: ack, sounds like an interesting idea. let's find some folks who are interested in this
16:30:45 <haypo> (sorry, i missed the "open discussion" :-))
16:31:15 <kzaitsev_mb> rpodolyaka1: so you can host the whole thing yourself
16:31:21 <dims> haypo: got a link to the email?
16:31:49 <dims> haypo: oh you want the poll right here, right now?
16:31:52 <haypo> http://lists.openstack.org/pipermail/openstack-dev/2015-October/077103.html
16:31:57 <haypo> [openstack-dev] [oslo] Require documenting changes with versionadded and versionchanged
16:32:09 <haypo> dims: right here, right now :)
16:32:10 <dims> haypo: we probably don't have quorum :) need dhellmann and harlowja and others
16:32:14 <kzaitsev_mb> dims: yep, I think I'll write an email, as soon as I get some prototype working =) Just to show some way I imagine it should work )
16:32:24 <dims> kzaitsev_mb: cool
16:32:43 <haypo> dims: it's just a way to discuss documentation :)
16:32:55 <kzaitsev_mb> dims: I was just thinking, that oslo.* is the right place to store common middleware/interfacing code.
16:32:59 <haypo> i don't expect that we take a decision right now ;)
16:33:00 <rpodolyaka1> kzaitsev_mb:  ah, ok. well, would be nice to see if there are any advantages over 'classic' ELK then ;)
16:33:19 <haypo> dhellmann announced that something will come making documenting changes easier
16:33:22 <kzaitsev_mb> so wanted to ask around if there are any strong objections or smth )
16:33:23 <dims> kzaitsev_mb: ack
16:33:34 <amrith> thanks dims I will follow up with dhellmann offline
16:34:03 <haypo> "Stand by for some announcements about release notes management coming next week that will help solve this problem in a different way." where next week = this week :)
16:34:05 <kzaitsev_mb> rpodolyaka1: what's 'classic' ELK? =) sorry for dumb questions =)
16:34:21 <dims> haypo: we have a session for you at Tokyo :) i'll add you as the chair
16:34:26 <rpodolyaka1> kzaitsev_mb: elasticsearch, logstash and kibana
16:34:30 <haypo> dims: oh, really?
16:34:34 <kzaitsev_mb> I see )
16:34:41 <haypo> dims: "let's document victor?"
16:34:47 <dims> haypo: http://mitakadesignsummit.sched.org/event/7ec1eacb794353e1fac18801b96e7e83
16:35:00 <dims> haypo: let me know if you are free at that time
16:35:15 <dims> "what makes haypo tick" :)
16:36:04 <haypo> dims: i will attend this session
16:36:08 <kzaitsev_mb> rpodolyaka1: well, it might be, that I'm not familiar with those =) and trying to reimplement the bicycle. but I believe, that it's still worth to try and build a quick prototype to show around, as I had lot's of positive experience with self-hosted sentry before +)
16:36:13 <dims> haypo: thanks
16:36:31 <dims> haypo: we'll need to start building etherpads with ideas
16:36:49 <rpodolyaka1> kzaitsev_mb: I agree with dims that ML thread is the better place to ask, you might want to send it to openstack-operators too
16:36:56 <dims> rpodolyaka1: ++
16:36:56 * krotscheck has an open discussion topic
16:37:01 <dims> krotscheck: go for it
16:37:09 <krotscheck> I'd like to create a library of common javascript things.
16:37:19 <krotscheck> I think calling it 'openstack/oslo-jslib' may make sense
16:37:27 <krotscheck> Is that something y'all would be willing to take on?
16:37:48 <kzaitsev_mb> rpodolyaka1: nice idea about the operators, will remember that )
16:38:10 <krotscheck> it would be seeded with code from here: http://git.openstack.org/cgit/openstack/ironic-webclient/tree/app/js/modules/openstack
16:38:11 <dims> krotscheck: i am willing to experiment if you can round up folks who can contribute as library cores
16:38:14 <haypo> dims: oh yes, i would help because the current description doesn't help me :-/
16:38:22 <dims> :)
16:38:36 <krotscheck> dims: Alllrightey.
16:38:52 <dims> thanks krotscheck
16:38:53 <krotscheck> dims: Is oslo the correct place for that, or would this make more sense as a non-python thing?
16:39:25 <dims> krotscheck: the big tent makes it easy to start anything, the os-win library is an example
16:39:38 <haypo> kozhukalov: cool, there are some unit tests :)
16:39:47 <dims> krotscheck: so the basic stumbling block is identifying others who are interested in this problem space
16:39:55 <dims> for reviews, ideas, etc
16:40:04 <krotscheck> dims: Yes, but what's the _right_ way to do this? [/german_obsessive_compulsive]
16:40:45 <dims> krotscheck: follow the pattern of os-win, round up some fellow contributors and there's a infra page on how to get a project started
16:41:04 <dims> krotscheck: http://docs.openstack.org/infra/manual/creators.html
16:41:37 <krotscheck> Gotcha
16:42:17 * dims checking if everyone got what they needed :)
16:42:35 <dims> last question, who will not make it to Tokyo?
16:42:48 <kgiusti> o/ :(
16:43:03 <dims> kgiusti: ow, we'll miss u :)
16:43:13 <jecarey> o/
16:43:14 <kgiusti> austin for shure!
16:43:32 <dims> kgiusti: need your ok on this one https://review.openstack.org/#/c/236168/
16:43:36 <dims> kgiusti: ++
16:43:55 <kgiusti> dims: will do.
16:43:57 <dims> hey jecarey, we were just wrapping up. any thing you want to add?
16:44:20 <jecarey> not going to Tokyo
16:44:40 <dims> jecarey: see you in austin then! :)
16:44:44 <dims> thanks everyone
16:44:51 <dims> #endmeeting