15:00:21 <rhochmuth> #startmeeting monasca
15:00:22 <openstack> Meeting started Wed Sep  7 15:00:21 2016 UTC and is due to finish in 60 minutes.  The chair is rhochmuth. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:23 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:24 <rhochmuth> o/
15:00:26 <openstack> The meeting name has been set to 'monasca'
15:00:27 <rbak> o/
15:00:30 <acabot> o/
15:00:39 <shinya_kwbt> o/
15:00:50 <koji> o/
15:00:51 <rhochmuth> We have a fairly verbose agenda today
15:00:57 <witek> o/
15:00:58 <rhochmuth> https://etherpad.openstack.org/p/monasca-team-meeting-agenda
15:01:08 <rhochmuth> Agenda for Wednesday September 7, 2016 (15:00 UTC)
15:01:08 <rhochmuth> 1.	Planning and Reviews
15:01:08 <rhochmuth> 1.	Cassandra/Monasca Summit
15:01:08 <rhochmuth> 1.	Monday, September 12, 8 AM MST
15:01:08 <rhochmuth> 2.	Hyper-V Support in the Monasca Agent
15:01:08 <rhochmuth> 1.	Thursday September 8, 1:00 PM MST
15:01:08 <rhochmuth> 3.	Notification Engine Plugin Review
15:01:09 <rhochmuth> 1.	Not scheduled
15:01:10 <rhochmuth> 4.	Events Design Planning and Review
15:01:10 <rhochmuth> 1.	Sometime next week
15:01:11 <rhochmuth> 2.	Sporadic metric status ?
15:01:11 <bklei> \o
15:01:11 <rhochmuth> 3.	ORM/Postgres - changes ready for closer review
15:01:25 <rhochmuth> Is it o/ \o
15:01:31 <rhochmuth> i forget
15:01:51 <witek> does it matter? :)
15:02:17 <rhochmuth> it might matter for listing the attendees for the meeting in the irc logs
15:02:22 <rhochmuth> but i don't know for sure
15:02:46 <rhochmuth> i first wanted to cover some upcoming events
15:02:54 <rhochmuth> #topic Planning and Reviews
15:03:08 <rhochmuth> Cassandra/Monasca Summit
15:03:26 <rhochmuth> i think I have everyone on an email invite that mentioned that they wanted to attend this
15:03:38 <rhochmuth> but I wanted to just list it here
15:04:16 <witek> did you get confirmation from Yuan or Andrei about Monday?
15:04:17 <rhochmuth> On Monday, September 12, 8 AM MST, we will have an on-line Skype/Lync session to discussion plans for the Cassandra work in Monasca
15:04:32 <rhochmuth> witek: yes i did
15:04:39 <witek> nice
15:04:56 <rhochmuth> Hyper-V Support in the Monasca Agent
15:05:09 <rhochmuth> There has been a blueprint and reviews in progress since May
15:05:30 <rhochmuth> The reviews are at, https://review.openstack.org/#/q/owner:cbelu%2540cloudbasesolutions.com+monasca+status:open
15:05:50 <rhochmuth> We'll be having an on-line session with cbelu tomorrow
15:06:00 <rhochmuth> Thursday September 8, 1:00 PM MST
15:06:09 <rhochmuth> If anyone wants to attend please let me know
15:06:26 <rhochmuth> This will be a design/code overview and walk-through
15:06:47 <rhochmuth> The goal would be to help us out understanding what was done to expedite the review process
15:07:24 <rhochmuth> I would like to have a on-line design/code review to discuss Notification Engine Plugins
15:07:51 <rhochmuth> There are three review in progress that add custom formatting for the hipchat/slack plugins as well as a new JIRA plugin
15:07:59 <rhochmuth> If anyone wants to be in on that please let me know
15:08:34 <rhochmuth> The idea is to help us understand the design and code quicker to get through the review process quicker
15:09:08 <rhochmuth> I haven't scheduled that review yet, so let me know if interested
15:09:12 <rhochmuth> email also works
15:09:38 <rhochmuth> Finally, we discussed the events api/design work last week
15:09:48 <rhochmuth> I haven't scheduled that meeting yet
15:10:02 <rhochmuth> i will try for late next week
15:10:17 <rhochmuth> and yes, i know all the fujitsu folks are interested in that
15:10:26 <witek> :)
15:11:20 <rhochmuth> Is there any other areas to cover for upcoming design/code review sessions/meetings
15:11:45 <rhochmuth> There is also prep for the Summit
15:12:26 <rhochmuth> #topic Sporadic metric status ?
15:12:40 <witek> it is from Tomasz
15:13:12 <witek> he wanted to ask if Igor or Erickson would like to take this over
15:13:22 <rhochmuth> i have to admit i haven't been looking at that review very closely
15:13:36 <rhochmuth> are igor or erickson here?
15:14:12 <witek> I'll tell Tomasz to write them an email
15:14:16 <rhochmuth> i think tomasz will need to contact directly
15:14:45 <rhochmuth> i'm certainly ok with it
15:14:59 <witek> ok, that's it I think
15:15:10 <rhochmuth> #topic ORM/Postgres - changes ready for closer review
15:15:29 <witek> Tomasz and Artur have been working on these recently
15:16:25 <witek> what is the status of devstack plugin on ubuntu-xenial?
15:16:47 <rhochmuth> shinya was workign on it
15:16:47 <witek> project-config want to move everyting to xenial
15:17:13 <shinya_kwbt> I'm working at https://review.openstack.org/#/c/366024/
15:17:27 <rhochmuth> thx shinya_kwbt
15:17:35 <shinya_kwbt> but still development
15:18:44 <witek> we have two options:
15:18:50 <rhochmuth> so, if you complete that, then that would allow tomasz and artur to integrate the postgres and orm plugin
15:19:48 <witek> rhochmuth: yes, otherwise we would have to convince project-config to accept tests on trusty
15:20:21 <rhochmuth> so is one option convince to accept on trusty, and option 2 to complete port to xenial?
15:21:01 <witek> yes, but I would switch the order
15:21:19 <rhochmuth> i was just listing them in any order
15:21:25 <witek> if the effort is not too big
15:21:47 <rhochmuth> shinya_kwbt: so do you have any time-frame you feel comfortable with
15:22:17 <shinya_kwbt> I will push migration but I don't know when it will be finished.
15:22:36 <rhochmuth> is there anything you need help with
15:23:01 <shinya_kwbt> if possible I want help.
15:23:31 <rhochmuth> do you have a list of specific areas?
15:23:49 <shinya_kwbt> there is some problem.
15:24:04 <shinya_kwbt> notifcation can't start.
15:24:33 <rhochmuth> very strange because that is one of the simpler components
15:24:35 <shinya_kwbt> and thresh cash.
15:24:39 <shinya_kwbt> crash
15:24:55 <rhochmuth> everytime thresh crashes you get cash
15:25:29 <rhochmuth> so, i'll look into some help on notification and thresh
15:25:42 <shinya_kwbt> I get cash? yes I pay cash? no
15:26:12 <shinya_kwbt> thank you.
15:26:43 <rhochmuth> it tomasz and artur can keep helping that would also be good
15:26:50 <witek> thanks shinya_kwbt and rhochmuth
15:26:54 <rhochmuth> the threshold engine crashing bother me
15:26:58 <shinya_kwbt> kafka api persister strom are working.
15:26:59 <witek> yes, sure
15:27:02 <rhochmuth> that sounds like a storm issue on xenial
15:28:07 <rhochmuth> So, I think that covers this topic
15:28:23 <rhochmuth> we'll try and get xenial working/completed
15:28:25 <shinya_kwbt> it seems thresh cash leeds read only file systems
15:28:29 <witek> the tests for ORM should be ready, so if we want to include then in newton we should decide for trusty or xenial soon
15:29:06 <rhochmuth> we've missed the date for getting them into newton
15:29:34 <witek> ok
15:29:36 <rhochmuth> although we could update the newton branch
15:29:46 <rhochmuth> if that was important
15:29:56 <witek> well, not really that much
15:30:01 <rhochmuth> but, newton is supposed to be critical bug fixes at this point
15:30:16 <witek> fair enough
15:30:32 <rhochmuth> #topic Dimensions API
15:30:39 <rhochmuth> https://review.openstack.org/#/c/350830/
15:30:50 <rhochmuth> We were discussing this last week
15:30:55 <rhochmuth> and via some emails on the side
15:31:16 <rhochmuth> my plan going forward is the following:
15:31:40 <rhochmuth> Return the dimensions names and values as a 1-D array in the "elements" field
15:31:50 <rhochmuth> each name and value is unique
15:31:56 <rhochmuth> therefore the returned list is unique
15:32:06 <rhochmuth> IDs are not necessary
15:32:23 <rhochmuth> Additionally, I'm wondering about the metric names resource
15:32:38 <rhochmuth> The documentation at, https://github.com/openstack/monasca-api/blob/master/docs/monasca-api-spec.md#response-examples-5
15:32:57 <rhochmuth> descirbes a 1-D array, simialr to how how i think dimension names and values returned
15:33:11 <rhochmuth> However, the doc doesn't exactly match the response in the code
15:33:27 <rhochmuth> the code returns metric IDs, which I dno't understand the usefullness
15:33:44 <rhochmuth> Anyway, any questions/issues with the above
15:33:58 <rhochmuth> I'm assuming rbak and bklei might have some
15:34:08 <rbak> Going back and changing the existing dimensions values will break Grafana
15:34:21 <rhochmuth> Yes, Grafana will need to be fixed
15:34:33 <bklei> it seems you feel strongly about it -- so whatever
15:34:51 <rbak> Alright, as long as everyone knows there will updates required.
15:35:30 <rhochmuth> rbak: how difficult are the changes to Grafana
15:35:47 <rbak> Not difficult, it just adds more complexity.
15:36:52 <rhochmuth> i would volunteer for making them, but i dont' know what i'm getting into
15:37:02 <rhochmuth> and my coding time has diminshed
15:37:12 <rhochmuth> lot's of reviews though
15:37:26 <rbak> I would need to for the new endpoint anyway, so it's not out of my way to do it.
15:37:44 <rhochmuth> i'll take that as you don't trust me
15:37:53 <rhochmuth> :-)
15:37:54 <witek> :)
15:38:03 <rhochmuth> thanks rbak
15:38:14 <rbak> If you want to give it a shot feel free, it just sounded like you didn't have time.
15:38:34 <rhochmuth> i probably don't
15:39:03 <rhochmuth> if you are going to making changes, which you have to anyway for the dimensions names, then i would prefer that you do it
15:39:19 <rhochmuth> it will take be a while to spin-up on grafana again
15:39:27 <rhochmuth> and have to learn a new programmign language too
15:40:05 <rbak> That's fine, I'll be able to get to it as soon as the api changes merge.
15:40:05 <rhochmuth> rbak: do you have a reason for the metric IDs
15:40:15 <rhochmuth> in the metric names resource
15:40:33 <rhochmuth> kaiyan pointed that out to me, but it doesn't match the doc
15:40:44 <rhochmuth> and i'm not sure if they are used anywhere
15:40:50 <rhochmuth> it doesnt' hurt keepign them
15:40:57 <rhochmuth> but i was wondering if they shoudl be removed
15:41:04 <rbak> Like I said before, the metric ids and 2d array were for consistency between responses.  It allows the client to treat and merge all pages the same regardless of the endpoint.
15:41:39 <rbak> It kept things clean and simple for clients, but it can be done without.
15:41:40 <rhochmuth> do you want to keepo them or remove them
15:42:17 <rbak> I would prefer that all of the endpoints work the same way, but it sounds like you already decided against that.
15:43:03 <rhochmuth> i don't see a compelling reason to wrap a 1-D array in a 2-D structure a list of list
15:43:57 <rbak> And that's fine.  I just wanted to throw out my views as someone writing a client for the api.
15:44:45 <rhochmuth> well, maybe i don't have an appreciation for the client-side completely, but i still think the 1-D structure is best
15:45:20 <rbak> That's fine.  In that case an id doesn't really do anything.
15:45:57 <rhochmuth> well, let me know if you think it could be removed and kaiyan can make that change too
15:46:17 <rhochmuth> she has got all the code opened up right now
15:46:23 <rbak> If we're going with 1d it can be removed.
15:46:35 <rbak> Same with the dimension values if we switch that to 1d
15:46:49 <rhochmuth> ok, thanks, i'll ask kaiyan to make those changes
15:47:05 <rbak> As long as we don't have a 2d array with no id it should work.
15:47:23 <rhochmuth> that is what i thought, so thanks for validating
15:47:29 <rhochmuth> i jsut wanted to be sure
15:48:18 <rhochmuth> #topic Add python multiple metrics for measurement and statistics
15:48:33 <rhochmuth> https://review.openstack.org/#/c/315742/
15:48:50 <rhochmuth> Rbrandt has that review finally passing gates and it shoudl be ready for review
15:49:34 <rbak> Is this just brining the python in line with the java group_by?
15:49:59 <rhochmuth> It is mainly about that, but there were some adjustments to the Java InfluxDB code
15:50:14 <rhochmuth> The Java code was mostly complete/done already
15:50:28 <rbak> But this still allows only group_by=*  ?
15:50:34 <rhochmuth> The Vertica code was not touched
15:51:41 <rhochmuth> Correct rbak
15:51:49 <rhochmuth> this is for group_by=* for the Python code
15:51:59 <rbak> Alright, just making sure I understood what I was looking at.  Thanks
15:52:08 <rhochmuth> group_by=dimension_name is another review
15:52:22 <rhochmuth> that has some merge conflicts
15:52:26 <rbak> yeah, I've been watching that one.
15:52:44 <rhochmuth> rbrandt is working on that one as we type
15:53:08 <rhochmuth> #topic Hyper-V support
15:53:17 <rhochmuth> https://review.openstack.org/#/q/owner:cbelu%2540cloudbasesolutions.com+monasca+status:open
15:53:29 <rhochmuth> It looks like these reviews are mostly ready to go
15:53:47 <rhochmuth> There will be a design/code review session on-line tomorrow if anyone is interested
15:53:58 <rhochmuth> If anyone wants to attend please let me know
15:54:38 <rhochmuth> #topic Monasca Notificaiton Engine plugins
15:54:46 <rhochmuth> 1.	Add Jira plugin: https://review.openstack.org/#/c/348001/
15:54:46 <rhochmuth> 2.	Add custom formatting: https://review.openstack.org/#/c/349097/
15:55:08 <rhochmuth> I'm trying to setup another on-line design/code review with haneef next week
15:55:31 <rhochmuth> again, mainly to expedite the review process, but in this case there might be some high-level discussions possibly on the jira plugin
15:55:45 <rhochmuth> let me know if interested and i can schedule with you included
15:56:01 <rhochmuth> #topic misc
15:56:07 <rhochmuth> We have five minutes left
15:56:14 <rhochmuth> are there any other topics
15:56:42 <rhochmuth> there are a bunch of other misc reviews and bug fixes
15:57:51 <rhochmuth> i guess that wraps up this session then
15:57:54 <rhochmuth> thanks everyone
15:58:00 <witek> thanks
15:58:13 <shinya_kwbt> thx
15:58:31 <bklei> thx roland
15:58:58 <rhochmuth> hasta la vista
15:59:04 <rhochmuth> #endmeeting