15:01:54 #startmeeting monasca 15:01:55 Meeting started Wed Feb 13 15:01:54 2019 UTC and is due to finish in 60 minutes. The chair is witek. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:55 hello 15:01:56 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:01:58 The meeting name has been set to 'monasca' 15:02:06 hi everyone 15:02:16 hello 15:02:29 hi 15:02:35 hi 15:02:42 the agenda for today as usual here: 15:02:47 https://etherpad.openstack.org/p/monasca-team-meeting-agenda 15:03:06 could we start with merging APIs? 15:03:19 sure 15:03:28 #topic merge APIs 15:03:32 Hi, regarding the Merge Monasca APIs. In the spec is written that the monasca-common dependency can be removed from the monasca-api. We are wondering where should end up the kafka-client code, since it is being used by other modules like monasca-persister. 15:04:13 Good point - would be nice to remove that kafka client code! 15:04:19 the spec says that API relevant code should be moved 15:05:16 if we remove Kafka code, we have to implement it in every component 15:05:26 or make a dependency to API 15:06:10 What about the work to remove the forked client? 15:06:27 I suppose in the longer term we will use an external client library? 15:06:44 I have picked up consumer implementation with confluent client again 15:07:12 still, some wrapper around it would be useful 15:07:49 I see, so the monasca-api will continue importing from monasca-common the kafka-client lib. 15:08:30 corresponding to the modules in https://github.com/openstack/monasca-common/blob/master/monasca_common/kafka/ 15:08:45 Ah - I had imagined that all the monasca components would just add the external Kafka library to their requirements. I didn't think about a wrapper. Could that be it's own library? 15:10:46 another alternative would be add this wrappers to individual components directly if it simplifies things 15:11:37 yes, i'd hope it'll be pretty light weight, so that could be simpler. 15:12:09 yes, new Kafka code should be simplier 15:12:40 do we have other shared code which would still live in monasca-common? 15:13:15 I don't think so 15:13:35 A yes... 15:13:43 the alarm expression parser 15:13:54 it is being used in monasca-ui as well 15:14:03 api and UI? 15:14:48 Thinking about adding the wrapper to each project, it scares me a little bit, given that there could be quite a few projects using it, it could easily diverge. 15:15:10 the code was copied from api to common to allow UI to used it 15:15:32 https://github.com/openstack/monasca-common/commit/086b9efcaf1c64693a2083dc3ef6115de7ae7b30 15:16:28 I'd say then, let's leave the other parts in monasca-common for now and finish API work first 15:16:49 Sounds sensible. 15:17:26 +1 15:17:34 +1 15:17:41 chaconpiza: ? 15:18:14 +1 15:18:21 :) nice, thanks 15:18:37 And thanks for doing the work :) 15:18:45 how is the progress by the way? 15:19:08 So far Adrian and me are extending the coverage for api and common 15:20:00 cool, please ping us for reviews 15:20:13 thanks 15:20:29 srwilkers: are you around? 15:20:40 hey witek 15:20:46 got a bit of a late start today 15:20:48 i'm around 15:20:49 hi 15:21:05 can we update shortly on OSH? 15:21:06 hello :) 15:21:09 sure 15:21:16 #topic openstack-helm 15:21:41 I have just copied the charts from monasca-helm repo to openstack-helm-addons 15:21:55 https://review.openstack.org/636597 15:22:26 nice - i saw that this morning 15:22:42 big... 15:22:46 wow, that is a lot of yaml 15:22:52 the idea is to adopt that code to OSH standards and eventually move to OSH repo 15:23:16 it is copied one to one from monasca-helm 15:24:23 srwilkers: could you point us how you would like the code to look like? 15:24:33 yep :) to get the linting jobs happy to get it ready for the initial merge, we'll need to mark the directories coming in with the chart as irrelevant files here: https://github.com/openstack/openstack-helm-addons/blob/master/.zuul.yaml#L16 and here: https://github.com/openstack/openstack-helm-addons/blob/master/.zuul.yaml#L24 15:24:56 sure, i'll attempt to without getting too much into the weeds on it 15:25:18 cool, thanks 15:25:30 i'll pick on glance, as it's my favorite one to pick on: https://github.com/openstack/openstack-helm/tree/master/glance/templates 15:26:05 you'll see we have various templates for things like the db-init job, the db-sync job, the ks-user job, etc 15:26:32 we'll need to massage the various charts coming in with monasca to adopt those types of standards. things like zookeeper and influxdb are obviously going to be different 15:27:07 but at least for the monasca components themselves, we'd want them to use the same sort of patterns and standards where applicable so the user experience of the chart matches the other openstack services 15:28:04 i need to revisit our documentation as well, as this was documented somewhere once upon a time. if it's not anymore, i'll work to get it updated to provide a clear footing to get things moving in the right direction 15:29:06 great, any help and pointers to docs very appreciated 15:29:10 im personally excited to get this rolling, as monitoring tools have a special place in my heart 15:29:18 :) 15:29:20 of course -- i'll give a shout here when i've got the doc situation sorted out 15:29:38 and feel free to ping me anytime for questions/help/pointers/whatever 15:29:42 that's our valentin's day present :) 15:30:14 it's a great one :D 15:30:25 i'm being summoned to stand up :) 15:30:55 :) 15:31:11 thanks srwilkers 15:31:51 also if anyone wants to help in that effort, very welcome 15:32:09 I'd like to, just bogged down with other things 15:32:44 zookeeper has been a thorn in my side this week. As such, I'll mention we should move the helm charts to 3.4.13 from 3.4.10 which I see in the values.yaml 15:33:25 srwilkers: one more question, I guess we can stay with our Alpine images, at least for the start, right? 15:34:39 joadavis: I think that should not be a problem, thanks for the pointer 15:35:49 we can clarify later, let's move on 15:35:54 #topic reviews 15:36:41 https://review.openstack.org/#/c/622361/ 15:37:24 looks ready for workflow approval 15:37:42 there are two things in this review: using monasca_db command in DevStack and fix in sub_alarm table definition 15:39:01 in Python 3 we had a problem with String type 15:39:33 https://review.openstack.org/#/c/634379 15:39:58 this one sets up a standard release notes job 15:40:52 it looked good, I just didn't workflow it to give others a chance to review (trying to stick to at least having two reviews before merging) 15:40:53 the output can be checked here: http://logs.openstack.org/79/634379/1/check/build-openstack-releasenotes/d4f9c11/html/ 15:41:19 https://review.openstack.org/#/c/634922/ 15:41:51 that's a trivial fix in default configuration value 15:42:47 https://review.openstack.org/#/c/633194/ 15:43:23 I have given -1 on that one because I'm not sure if it works correctly 15:43:54 glad you tested that 15:43:59 the enumeration values are still initialized with sql script: email, webhook and pagerduty 15:44:18 on the other hand, in notification configuration all plugins are enabled 15:44:27 per default 15:45:05 there is logic which should sync these during the start of the service 15:45:36 for some reason it doesn't work, but I'm not sure why 15:47:16 there are many more reviews from Thomas: 15:47:18 https://review.openstack.org/#/q/projects:openstack/monasca+is:open+owner:%22Thomas+Bechtold%22 15:48:15 mainly clean-up for oslo.config 15:48:44 https://review.openstack.org/#/c/636150/ 15:48:58 Fix alarm-definition-list in Python 3 15:49:25 thanks dougsz 15:49:43 np 15:49:56 did you look more at client code in terms of Python 3 support? 15:50:04 I ran away 15:50:09 haha 15:50:21 I'm sure it won't take too long, just need to find the time. 15:51:16 dougsz: chaconpiza: do you think it's something where piotrowskim could help? 15:51:29 Merged openstack/monasca-notification master: Configure releasenotes job https://review.openstack.org/634379 15:51:52 sorry, I meant Dobroslaw, not dougsz 15:52:02 jep, could be another task for him 15:52:03 There are more small task related to the hacking lib 15:52:31 yes, several repos could use a hacking lib review 15:52:34 https://storyboard.openstack.org/#!/story/2004930 15:53:40 the last two reviews from the list: 15:53:45 https://review.openstack.org/#/c/636150/ 15:53:49 ups 15:53:56 https://review.openstack.org/#/q/status:open++branch:master+topic:policy-names 15:54:43 these ones are about updating the policy names to OpenStack conventions 15:55:07 the reviews have been up for a while now already 15:55:45 some more reviews? 15:56:41 #topic open stage 15:56:48 Just short info about https://github.com/monasca/monasca-docker 15:56:48 Looks like it's no longer possible to build monasca-api 15:56:48 Problem is that it's using really old PR for building https://review.openstack.org/#/c/417163 15:56:48 and newer version of prometheus_client is incompatible anymore with this code 15:56:48 I'm trying to push changes to build Docker in zuul but zuul don't like me... 15:58:40 shouldn't we use new master branch code? 15:59:24 Merged openstack/monasca-notification master: Replace notification.conf with monasca-notification.conf https://review.openstack.org/633163 15:59:55 the thing is this code is used to push some metrics to monasca-sidecar and I never properly looked into what it used for... 15:59:58 I have to jump to another meeting now 16:00:06 thank you everyone for joining 16:00:12 thank you 16:00:13 and for the productive meeting 16:00:13 thanks 16:00:16 thanks 16:00:17 thanks all 16:00:18 bye 16:00:22 bye 16:00:25 thanks, bye 16:00:25 #endmeeting