15:00:17 <rhochmuth> #startmeeting monasca
15:00:18 <openstack> Meeting started Wed Jan  6 15:00:17 2016 UTC and is due to finish in 60 minutes.  The chair is rhochmuth. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:22 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:24 <openstack> The meeting name has been set to 'monasca'
15:00:33 <rhochmuth> o/
15:00:36 <ddieterly> o/
15:00:47 <fabiog> Happy New Year!
15:00:51 <rbak> o/
15:00:52 <rhochmuth> welcome back everyone
15:00:57 <bklei> o/
15:01:00 <shinya_kwbt> o/
15:01:11 <bklei> happy new year!
15:01:15 <rhochmuth> agenda has been posted at, https://etherpad.openstack.org/p/monasca-team-meeting-agenda
15:01:25 <rhochmuth> Agenda for Wednesday January 6, 2016 (15:00 UTC)
15:01:25 <rhochmuth> 1. Mid-Cycle/Meet-up go/no go and proposed agenda (fabiog)
15:01:25 <rhochmuth> 2. What e-mail validator do we use at python monasca-api? (shinya)
15:01:25 <rhochmuth> 2.1 Voluptuous validator which is used by monsaca-api has no e-mail validator but has url one.
15:01:25 <rhochmuth> 2.2 lepl.apps.rfc3696 - MPL 1.1 License http://www.acooke.org/lepl/
15:01:25 <rhochmuth> 2.3 validate_email - LGPL License https://pypi.python.org/pypi/validate_email
15:01:26 <rhochmuth> 2.4 Regex
15:01:26 <rhochmuth> 3. TWC updates
15:01:27 <rhochmuth> 3.1 UI update
15:01:27 <rhochmuth> 3.2 Any progress on caching from HPE?
15:01:28 <rhochmuth> 3.3 Openstack requirements in 3 files, required for rally tests
15:02:06 <rhochmuth> Also, we are a bit backed up on reviews for a number of reasons
15:02:42 <rhochmuth> Mainly, the holidays and the gates are not functioning right now due to changes in the tempest tests
15:02:59 <rhochmuth> So, let's start with the mid-cycle meet-up
15:03:09 <rhochmuth> #topic mid-cycle
15:03:20 <rhochmuth> fabiog, this is yours
15:03:23 <fabiog> here it is the doodle: http://doodle.com/poll/yy4unhffy7hi3x67
15:03:36 <fabiog> I have 6 people confirmed for Thu 28/29
15:03:59 <fabiog> and on the 28 we will share the day with Congress
15:04:14 <fabiog> meaning fostering integration and collaboration
15:04:48 <rhochmuth> i thought the congress discussion was going to be on tuesday or wednesday
15:04:56 <rhochmuth> leaving thursday and friday open for monasca
15:05:01 <rhochmuth> is that not correct
15:05:06 <fabiog> nope
15:05:15 <fabiog> Congress is 3 days
15:05:19 <fabiog> Tue to Thu
15:05:42 <rhochmuth> but i tought the monasca topics were going to be on tues/wed originally
15:05:51 <fabiog> yes
15:05:59 <fabiog> that was the original plan
15:06:03 <fabiog> but I am hosting both
15:06:13 <fabiog> and I cannot be in two places at the same time
15:06:15 <fabiog> :-)
15:06:27 <rhochmuth> ohh, i see
15:06:33 <fabiog> I thought that since you guys were coming, we could have a joint session on Thu
15:06:48 <rhochmuth> i thought you were only attending the monasca related discussions for congress
15:07:10 <fabiog> no, I am hosting the mid-cycle
15:07:15 <rhochmuth> so, i can attend the congress related discussions
15:07:21 <fabiog> yes
15:07:26 <rhochmuth> but, i'm concerned now about short-changing monasca
15:08:26 <rhochmuth> if folks are traveling for a monasca mid-cycle, and thursday is still on congress mainly, that leaves only one day for monasca discussions
15:09:23 <rhochmuth> i don't think most of the folks traveling are as interested in congress related discussions
15:09:41 <fabiog> I can get a second room and have the congress stuff running only in the morning
15:09:51 <fabiog> but I will be able to attend either one or the other session
15:10:00 <fabiog> I think this is doable
15:10:14 <rhochmuth> ok
15:10:27 <rhochmuth> so let's discuss other topics for the mid-cycle then
15:10:37 <rhochmuth> meant agenda
15:11:05 <rhochmuth> what else should be on the agenda for the mid-cycle, in addition to congress discussions
15:12:06 <rhochmuth> the logging api would be one area that i'm interested, but we don't have representation from fujitsu
15:12:14 <rhochmuth> at the mid-cycle
15:12:41 <bklei> i'd like to add an arch discussion about concurrency, and possible solutions
15:12:49 <bklei> concurrent queries, that is
15:13:19 <bklei> we're hitting some hard limits with vertica, not sure if influxdb, cassandra would be better
15:13:34 <rhochmuth> just a litle more context then
15:14:18 <bklei> we're working closely with the vertica team, but we're a couple orders of magnitude off from the concurrency we'd like to support
15:14:32 <bklei> it may be vertica isn't a great fit for our use case
15:14:48 <bklei> like 100 users, all loading a grafana dashboard at the same time :)
15:15:00 <bklei> and each dashboard translates to 150 queries
15:15:28 <fabiog> so you need to support 1000+ concurrent queries
15:15:30 <bklei> don't need to discuss here, just would like as a topic at the meetup
15:15:33 <bklei> like 20K
15:15:45 <bklei> ideally
15:15:47 <rhochmuth> ok, thanks for the context
15:15:56 <rhochmuth> i'm not sure about cassandra or influxdb either
15:16:17 <rhochmuth> personllay, i think we need to start considering adding an in-memory time-series database/cache
15:16:45 <bklei> right, something
15:16:59 <rhochmuth> ok, other topics then for the mid-cycle
15:18:23 <bklei> maybe an overview/demo of grafana 2.0?
15:18:45 <rhochmuth> has ryan completed that?
15:18:52 <bklei> update coming :)
15:19:16 <rbak> I should have it done by the mid-cycle
15:19:28 <rhochmuth> ok
15:19:56 <rhochmuth> there are other features that we could discuss, such as retention periods
15:20:01 <fabiog> I am wondering if we want to do a couple of days using webex instead and have people from Europe to attend too
15:20:02 <bklei> oh yes
15:20:17 <bklei> oh yes on retention periods rhochmuth
15:21:01 <rhochmuth> it does seem like the agenda is a bit limited
15:21:12 <fabiog> rhochmuth: that is my feeling too
15:21:31 <fabiog> rhochmuth: but we can have a couple of good webex sessions to iron out the issues and see the demos
15:21:54 <rhochmuth> i think i would push for a webex too, at this point
15:21:55 <fabiog> and if people in the US are willing to travel anyway we can be in a room with telepresence
15:21:57 <tgraichen> i think this would be very good - maybe at least one or a few
15:22:24 <rhochmuth> then i would also suggest that we have the monasca mid-cycle the following week
15:22:34 <fabiog> rhochmuth: yes
15:22:35 <rhochmuth> the first week of february sometime
15:22:39 <fabiog> sure
15:22:43 <rhochmuth> so as not to overlap with congress
15:22:51 <rhochmuth> i would like to attend the congress related discussion
15:22:57 <fabiog> what about two mornings
15:23:07 <fabiog> Wed and Thu?
15:23:14 <rhochmuth> yes, that would work well for me
15:23:21 <rhochmuth> mornings are best for folks in europe too
15:23:24 <fabiog> people already have time on wed this time
15:23:33 <fabiog> so we could start the webex at 7am to noon
15:23:44 <fabiog> so it is kind of feasible for Europe to follow
15:23:50 <rhochmuth> sure, that would work
15:23:56 <fabiog> and if needed we will repeat the next day
15:24:10 <fabiog> ok, then is set
15:24:29 <fabiog> I will send an invite to people and post a note in the monasca dev list
15:24:36 <rhochmuth> wed/thurs feb 3rd and 4th
15:24:40 <fabiog> yes
15:24:45 <rhochmuth> 7-12 PST
15:25:40 <rhochmuth> ok, sounds like we should move on to the next topic
15:25:58 <rhochmuth> #topic e-mail validator
15:26:14 <shinya_kwbt> Hello. this is mine.
15:26:23 <rhochmuth> hi Shinya
15:26:40 <shinya_kwbt> I'm writing python monasca-api notification methods e-mail validator.
15:27:03 <shinya_kwbt> Java monasca-api uses apache commons validator and javax.mail validator under review.
15:27:11 <rhochmuth> yes, i recall there were some problems with this
15:27:50 <shinya_kwbt> I see. And python momnasca-api uses voluptuous validator. But voluptuous has no e-mail validator but has url one.
15:28:15 <rhochmuth> so, i wouldn't restrict to using voluptuous
15:28:28 <rhochmuth> we've had some disucssion on removing voluptuous
15:28:45 <rhochmuth> and hand-coding
15:29:55 <shinya_kwbt> Oh. I see.
15:30:41 <rhochmuth> their aren't any immediate plans to remove voluptuous, but if it was in your way
15:30:48 <rhochmuth> then i would be fine not using it
15:32:09 <shinya_kwbt> Is hand-coding for example regex better? or use other library?
15:32:36 <rhochmuth> i don't have experience beyond voluptuous
15:32:44 <rhochmuth> so, i can't answer that question
15:33:01 <rhochmuth> but, i think a regex is ok
15:33:04 <rhochmuth> in this case
15:33:51 <shinya_kwbt> O.K. I try to write with regex. thank you.
15:34:05 <rhochmuth> ok, thanks
15:34:11 <rhochmuth> there should be lot's of examples
15:34:32 <shinya_kwbt> yes.
15:34:49 <rhochmuth> ok, i think we are ready for the next topic
15:35:05 <rhochmuth> #topic ui update
15:35:10 <rbak> I guess that's me
15:35:15 <rhochmuth> it is
15:35:34 <rbak> I'm working on the grafana/keystone integration, and I have it partially working.
15:36:08 <rhochmuth> what was your overall design
15:36:09 <rbak> So you can log into grafana with keystone creds and it will assign you an org corresponding to your tenant, which shares dashboards.
15:37:13 <rbak> This isn't working yet, but the idea is that it will then use your creds to get tokens for the monasca queries so you don't need a token in the datasource.
15:38:05 <rhochmuth> so, will you login to horizon and grafana?
15:38:40 <rbak> I'd like to have horizon pass the necessary information to grafana, but I haven't looked at that yet.
15:38:41 <tgraichen> do you have keystone v3 support in mind from the beginning?
15:39:12 <rbak> No, I haven't thought about v3 support much, but I can look into tha.
15:39:44 <rhochmuth> tgraichen, this would probably be a good area for you to get input to rbak on
15:39:55 <tgraichen> ok - let me know if you need help or support testing v3 related stuff
15:40:01 <rbak> will do
15:40:17 <rhochmuth> rbak, is your latest code in github
15:40:32 <rbak> Not yet, I'll put it up once it's tested
15:40:40 <rbak> But fair warning, this isn't pluggable
15:40:55 <rhochmuth> i translated that into this isn't easy
15:41:02 <rbak> Until it's merged upstream you'll have to build your own grafana.
15:41:43 <rhochmuth> how difficult is it to get going with grafana 3
15:42:25 <rhochmuth> for example, if you clone the grafana 3 repo, clone your branch when avaialble
15:42:40 <rhochmuth> then do some configuration
15:42:55 <rbak> There's a grafana 3 now?
15:43:03 <tgraichen> i think he mans 2 :)
15:43:08 <rbak> Oh, good
15:43:17 <rhochmuth> it's the new year
15:43:20 <rbak> That had me worried.
15:43:37 <tgraichen> i think rbaks point was, that you can't use prebuilt grafana binaries anymore
15:43:43 <rbak> Right
15:43:57 <rbak> For the datasource you can just install and then drop the files in.
15:44:10 <rbak> For the keystone integration you have to clone and build.
15:44:17 <rhochmuth> got it
15:44:24 <rbak> It's not terrible, just some extra work
15:44:24 <rhochmuth> grafana 1 was javascipt only
15:44:29 <rhochmuth> so no compile step
15:44:40 <rbak> This work is primarily in go
15:44:57 <tgraichen> did you already contact the grafana maintainer, if he would be willing to include this upstream when ready?
15:45:26 <rbak> Not yet.  I wanted to make sure I knew this was possible to implement first.
15:45:35 <tgraichen> sure
15:45:37 <rbak> I've started looking into that though.
15:46:00 <tgraichen> maybe creating an issue in the grafana github when ready
15:46:33 <rbak> They seem to suggest just creating a pull request.
15:46:55 <tgraichen> ok
15:47:33 <rbak> That's all  I have unless there's more questions.
15:47:40 <rhochmuth> thanks rbak
15:47:54 <rhochmuth> looking forward to this
15:48:05 <rhochmuth> #topic caching
15:48:09 <rhochmuth> no updates
15:48:16 <rhochmuth> i just got back from the break
15:48:17 <bklei> i figured, with the holiday
15:48:24 <bklei> on your list?
15:48:42 <rhochmuth> i'm not sure
15:48:55 <rhochmuth> i've got a lot on my plate
15:49:24 <bklei> ok, maybe it'll be part of the mid-cycle discussion
15:49:39 <bklei> i don't think i'll get to it between now and then either
15:49:54 <rhochmuth> sounds like everyone is busy
15:50:00 <bklei> for sure
15:50:17 <rhochmuth> so, table this for the mid-cycle
15:50:26 <rhochmuth> along with concurrrent queries
15:50:31 <bklei> sounds good
15:50:50 <rhochmuth> #topic openstack requirements in 3 files
15:50:53 <bklei> that's me
15:51:13 <bklei> yan is working on rally tests, currently has an issue with incompatible requirements
15:51:25 <bklei> Incompatible requirement found; see http://docs.openstack.org/developer/requirements/
15:51:42 <bklei> she's saying we need to change openstack/requirements repo in global-requirements.txt, projects.txt, and upper-constraints.txt
15:51:49 <bklei> i guess it's in one, not all 3
15:52:15 <bklei> i don't know much about this, but any objections to us just adding monasca client to all 3?
15:52:45 <rhochmuth> fabiog added this recently, i guess in just the one spot
15:52:53 <bklei> Requirement set([Requirement(package=u'python-monascaclient', location='', specifiers='>=1.0.24', markers=u'', comment='', extras=frozenset([]))]) not in openstack/requirements
15:52:59 <rhochmuth> i don't see any issues adding to all 3 spot
15:53:00 <bklei> is her error
15:53:02 <bklei> ok
15:53:05 <fabiog> rhochmuth: bklei yes I added to global
15:53:33 <bklei> ok, if you don't add to the other two, either yan or i will
15:53:35 <fabiog> rhochmuth: but I was waiting for the change of oslo library
15:54:01 <rhochmuth> hmmm, i thought we got the changes for the oslo libraries compleete
15:54:06 <fabiog> bklei: the client is using an old oslo library that needs to be updated
15:54:18 <bklei> oh
15:54:21 <fabiog> once it is updated we can merge the global and the others
15:54:31 <fabiog> bklei: that is why is still not merged
15:54:34 <bklei> who's working that?
15:54:40 <rhochmuth> joe keen
15:54:55 <bklei> ok, can follow up with him and watch for changes
15:55:09 <fabiog> bklei: please keep me in the loop
15:55:11 <bklei> ok
15:55:18 <rhochmuth> i thought he had completed prior to xmas
15:55:30 <rhochmuth> is this just an issue for tagging again
15:55:37 <rhochmuth> do, i just need to apply a tag
15:55:58 <fabiog> these are the two related patches
15:56:00 <fabiog> https://review.openstack.org/#/c/251674/
15:56:14 <fabiog> https://review.openstack.org/#/c/252527/
15:56:43 <rhochmuth> i'll check with joe
15:56:53 <bklei> thx
15:57:22 <ddieterly> rhochmuth: looking for a +2 on https://review.openstack.org/#/c/254403/ and https://review.openstack.org/#/c/254425/
15:57:59 <rhochmuth> ok
15:58:06 <rhochmuth> but i was waiting on the gates working again
15:58:17 <ddieterly> the gate is working again
15:58:39 <ddieterly> i pushed the changes thru this morning after working with ghansyam last night
15:58:50 <rhochmuth> ok, thanks
15:59:05 <ddieterly> https://review.openstack.org/#/c/263529/
15:59:08 <rhochmuth> i'll +2 the other changes and start catching up again
15:59:11 <rhochmuth> on other reviews
15:59:32 <rhochmuth> we'll need to coveer cassandara next week
15:59:35 <bklei> i think we're out of time, can hold my cassandra question
15:59:37 <bklei> yup
15:59:47 <rhochmuth> bye everyone
15:59:48 <rhochmuth> thanks
15:59:50 <bklei> thx
15:59:52 <ddieterly> ciao
16:00:10 <rhochmuth> #endmeeting