15:00:02 #startmeeting monasca 15:00:03 Meeting started Wed May 4 15:00:02 2016 UTC and is due to finish in 60 minutes. The chair is rhochmuth. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:04 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:07 The meeting name has been set to 'monasca' 15:00:08 o/ 15:00:11 o/ 15:00:14 o/ 15:00:18 https://etherpad.openstack.org/p/monasca-team-meeting-agenda 15:00:31 hi everyone 15:00:37 hello 15:00:52 hi 15:00:56 it looks like we have a full agenda 15:00:57 hi 15:01:09 o/ 15:01:10 hi jayahn 15:01:14 hello 15:01:40 good to see you here :) 15:01:48 is everyone recovered from last week 15:01:53 i'm not 15:01:56 :) 15:02:13 so, there is an agenda that i posted the link for up above 15:02:22 the first topic is mic-cycle planning 15:02:27 #topic mid-cycle 15:02:56 we would be interested in hosting mid-cycle in Munich 15:03:05 at the end of July 15:03:15 i put in a travel request with my mgmt 15:03:26 but, haven't heard back yet 15:03:32 i only asked last night 15:03:49 I can do it in Munich only 07/27-29 15:03:57 i still need to get permission 15:04:06 hopefully, i'll know more in a couple of days 15:04:23 so, July 27, 28 and 29 15:04:25 will others be able to travel to munich? 15:04:35 who cares 15:04:44 :) 15:04:47 rhochmuth: don't be mean 15:04:58 i will not be able to travel to there.. almost for sure. 15:05:02 sorry, sometimes my fingers type on there own 15:05:18 alternatively Cisco could host it in SJC July 19 to 21 15:05:26 so, we'll need to get a head-count on who can travel 15:05:30 and see if this makes sense 15:05:45 we can set a doodle for that 15:05:48 let's try to get travel requests done by next week 15:06:00 ok, that sounds good too 15:06:18 jayahn, i don't recognoize your irc 15:06:40 can you introduce yourself 15:06:56 ah. yeap. 15:06:59 this is Jaesuk Ahn from skt. 15:07:11 from Korea, and it is midnight here. :) 15:07:12 ohh, got it 15:07:31 thanks for attending 15:07:39 no problem. 15:07:54 we are also talking about a new time slot for meetings 15:08:06 so, witek, you are going to create a doodle 15:08:11 that will be great 15:08:19 yes, for mid-cycle 15:08:20 i think i'm available for travel anytime 15:08:26 just don't know about travel budget 15:08:29 #action Witek sets doodle for mid-cycle 15:09:12 ok, let's sync-up next week on this topic and see if we can do a mid-cycle 15:09:13 for meeting time, midnight is not that bad for me. we can keep the current time. 15:10:08 I got the doodle up 15:10:13 :) 15:10:26 witek: sorry I was doing it without paying attention to the chat 15:10:28 http://doodle.com/poll/x6bf8ev6wvenfkdy 15:10:31 fabiog is a doodle master 15:10:32 np 15:10:54 rhochmuth: I should change my biz title into that :-) 15:11:30 Fabiog, Doodle Master Extrodianaire 15:11:34 at your service 15:11:47 Let's move on 15:11:57 #topic stable branch 15:12:06 https://review.openstack.org/#/c/306303/ 15:12:23 yes, we have that obvious bug in monasca-ui 15:12:49 in my opinion we should mergre into stable 15:13:12 i'm not sure about that -- i think we use /static at twc 15:14:17 hm 15:14:44 will confirm and add a comment to review 15:14:50 OK, thanks 15:15:35 thanks bklei 15:15:49 could you set STATIC_URL to / and have it work witek? 15:15:49 #topic reviews 15:16:26 but this would change the location of other static resources, right? 15:16:47 yeah -- i guess so 15:17:28 we can move on - will confirm if this patch would work for us 15:17:58 so there is revice https://review.openstack.org/#/c/307963/ 15:18:10 fix metric-list limits 15:18:14 twc REALLY wants this 15:18:29 looks like it's in good shape? 15:18:43 ok, i'll review today, and probably end up merging 15:19:04 awesome 15:19:15 witek, are you ok with this one 15:19:41 I haven't reviewed :( 15:19:57 it is basically a bug fix on metrics list 15:20:17 So, there is, https://review.openstack.org/#/c/289675/ 15:20:29 i started reviewing this yesterday 15:20:57 bklei, it looks like you've +1'd that one 15:21:08 have you done any performance testing at twc 15:21:26 just functional testing, no perf tests yet 15:21:28 also, to see the benefits of this change, you'll need to modify grafana 2 15:21:39 or rbak will 15:21:39 yeah -- rbak is eager to do that :) 15:22:03 ok, assuming i get time today, i'll do a review and merge 15:22:13 this shouldn't impact anyone 15:22:29 great -- yeah, shouldn't break any existing code path 15:22:32 the new capability is only enable if the group_by is specify as a qwuery param 15:22:49 so, this is low-risk 15:23:07 i'm assuming it will be merged today 15:23:21 https://review.openstack.org/#/c/306621/ 15:23:35 that's me 15:23:48 i think i've rolled in feedback to date 15:24:04 well, you didn't get my feedback 15:24:10 :-) 15:24:10 any objection to this given the larger conversation with the neutron ptl? 15:24:13 :) 15:24:28 i'll start reviewing this one too 15:24:45 we didnt' discuss this with armando 15:24:51 but, we could add him to the review 15:25:02 cool. i'm guessing if we get neutron to start pushing data to monasca, it will give us more than this. but this is the bare minimum our customers are asking for 15:25:10 sure, will add him 15:25:28 but i'm guessing, based on the conversations with him -- real code is years away :) 15:25:59 is it armando dominguez? 15:26:45 Armando Migliaccio 15:26:50 thx 15:27:04 he was at the 4th session last Wednesday 15:27:06 added him 15:27:11 did you meet him 15:27:25 i was in that session, but didn't talk with him directly 15:27:52 i did talk with maryam tahhan -- she was thinking about doing something similar to my patch, so i added her 15:27:52 i'll add some folks from the hpe neutron team as well 15:27:59 cool 15:28:00 like ed bak, ryan's dad 15:28:05 +1 15:28:17 small world 15:28:25 hi syd 15:28:31 morning 15:28:45 sick as a dog, caught something on the flight back 15:28:50 yes, ed and ryan might be the first father son team in openstack 15:28:59 that's awesome 15:29:12 we might have an entire generation of openstackers 15:29:15 next 15:29:41 https://review.openstack.org/#/c/286281/ 15:29:49 the infamous vertica hint 15:29:56 ok, will review 15:29:57 :) that's me too 15:30:08 thx. turned out to be pretty small change 15:30:19 and keeps existing behavior, unless the cfg setting is there 15:30:45 btw, what happened with removing the temp tables 15:30:59 isn't there code that needs to be submitted for that one too 15:31:31 yes, i'm on the hook for that 15:31:38 hope to get to that next week 15:31:41 ok, sounds good 15:31:49 are you using that change in prodiuction already? 15:31:50 i'd tested it, but needed cleanup. it works though 15:32:15 no -- not using the change yet 15:32:27 ok, will wait for your review 15:32:29 so far, we haven't carried anything local 15:32:32 cool 15:32:37 we're going to be picking up some of these changes too 15:32:50 yeah -- that one in particular should really help vertica 15:32:58 KV -- we'll see 15:33:11 https://review.openstack.org/#/c/251674/ 15:33:24 https://review.openstack.org/#/c/285000 15:33:26 that one and the next go together 15:33:39 Joe Keen is working on that 15:33:54 yeah -- until this gets fixed, the rally tests we're trying to push for monasca are on hold 15:34:06 Ahhh, I see 15:34:09 python client needs to be in global reqs for rally tests to work with monasca 15:34:24 So, I'm assumign Joe will try and get this resolved fairly soon 15:34:33 awesome 15:34:45 bklei, do you know how much of Monasca needs to be in the projects.txt for your tests to work? 15:34:48 shoudl probably reach out to him if it starts getting delayed again 15:35:31 jkeen -- not sure, i can connect you with yan, who knows more than me (via email) 15:35:54 That would be great. 15:36:12 thx 15:36:52 so while we have joe and michael i wanted to let folks know about periodic notifications 15:37:00 #topic periodic notifications 15:37:51 joe, can you post link to blueprint? 15:39:03 https://blueprints.launchpad.net/monasca/+spec/heat-integration 15:39:52 https://review.openstack.org/#/c/296813/ 15:40:01 basically we are adding in the ability to send a notfiication over and over at a given period until the state of the alarm it is associated with changes 15:40:07 https://review.openstack.org/#/c/310534/ 15:40:20 https://review.openstack.org/#/c/308502/ 15:40:23 we are only allowing this on webhooks, we are doing this for heat autoscaling 15:40:26 so we're saying 0 or 1 -- but should it be 0 or n? 15:40:34 for now 0 or 1 15:40:39 0 or 1 makes the problem much easier for now. 15:40:44 but going forward we left it open to add more intervals 15:40:49 We can expand that later to a larger predefined set of intervals. 15:40:53 pretty easy to add more intervals 15:40:57 ok, and heat works with 1 minute re-alarms? 15:41:07 do we want units in minutes or seconds 15:41:23 like -- is that what ceilometer does? 15:41:30 Yes, heat requested that timing. 15:41:35 perfect 15:42:07 so, i have a couple of comments i'll add to the reviews 15:42:18 this does involve a schema change to mysql 15:42:24 postres, … 15:42:30 right? 15:42:41 yes and also adding another kafka topic 15:43:13 oooh, we'll need to track that in our puppet module 15:43:20 yup 15:43:37 could you add the topic details to the blueprint? 15:43:48 or is it in the docs in one of your reviews 15:44:04 I can add more information to the blueprint 15:44:17 gracias 15:44:29 basically schema changes for databases and kafka changes 15:45:04 ok, hopefully we'll have an upgrade path? 15:45:16 what's an upgrade path? 15:45:43 bklei: schema update script? 15:45:45 upgrading the mysql schema 15:45:47 yeah 15:46:10 (we can't drop the db and start over) 15:46:10 we don't do db migrations, if that is what you mean 15:46:16 yup 15:46:27 i think it is just the addition of a column 15:46:37 ok, so an alter table command 15:46:39 https://review.openstack.org/#/c/308502/13/devstack/files/schema/mon_mysql.sql 15:46:57 currently called periodic_interval at the link above 15:47:02 not too bad, we can migrate that ourselves 15:47:08 yes, just alter table 15:47:32 possible name change on the column based on feedback 15:48:02 ok, we'll just need to remember to add the column when we pull that in 15:48:03 thx 15:48:41 ok, will close on that topic then and look for feedback on reviews 15:48:59 i'm guessing this is a week or two from merging 15:49:53 on a side-note, we are looking into dropping the mysql from the api, and just have sqla 15:50:13 also, looking into the notification engine 15:51:00 #topic vitrage 15:51:17 so we had some interesting discussions with vitrage at the summit 15:51:24 they would like to integrate with monasca 15:51:34 i posted some requirements that they listed 15:51:51 where? 15:51:59 https://etherpad.openstack.org/p/monasca-team-meeting-agenda 15:52:07 ah, ok 15:52:17 i don;'t think we'll have time to discuss today unfortunately 15:52:25 but, i wanted to tee that one up 15:53:06 the alarm CRUD operations seems like good things to have 15:53:54 yeah, i think that makes sense, but i think it is more complicated than oru early discussions led us to believe 15:54:10 so, we'll need to have some design sessions on this one 15:54:21 on this topic 15:54:54 witek: does everythign make sense for the sproadic metrics and the latched alarms 15:55:32 yes, Tomasz rewrites his change in monasca-thresh 15:55:51 ok, hopefully we got it this time around 15:56:38 he has put what he's done sofar to gerrit 15:56:47 comments are wellcome :) 15:57:02 ok, i'll start taking a look 15:57:08 and i'll ping craig 15:57:34 so, we are down to two minutes 15:57:46 and we didnt' cover everything 15:58:05 also, next week the transform engine will be avaialble for review 15:58:14 cool 15:58:16 maybe we need two meetings 15:58:21 a week 15:58:22 cool 15:58:35 :( 15:58:43 there is a blueprint on the transform engine 15:59:04 but, the actual design is still in progress of getting posted to the monasca wiki at 15:59:31 https://wiki.openstack.org/wiki/Monasca#Transform_and_Aggregation_Engine 15:59:38 Actually, it looks like it is all there 15:59:44 so, please start taking a look 15:59:52 they just published that last night 16:00:12 we've run out of time again 16:00:29 thank you Roland 16:00:30 i need to close the meeting 16:00:34 thanks 16:00:37 thanks Roland 16:00:37 bye everyone 16:00:38 cya 16:00:52 #endmeeting