17:00:14 <gibi> #startmeeting nova notification
17:00:14 <openstack> Meeting started Tue Apr 24 17:00:14 2018 UTC and is due to finish in 60 minutes.  The chair is gibi. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:16 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:18 <openstack> The meeting name has been set to 'nova_notification'
17:01:24 <mriedem> o/
17:01:35 <gibi> o/
17:01:51 <gibi> I have only one thing
17:02:17 <gibi> we kept the bp https://blueprints.launchpad.net/nova/+spec/add-server-group-remove-member-notifications in Pending Approval state
17:02:28 <gibi> to look at the implementation first
17:02:47 <gibi> and Takashi proposed the impl https://review.openstack.org/#/c/559076/
17:03:04 <gibi> it is on my list to look at
17:03:17 <gibi> but I wanted to have a generic reminder
17:03:53 <gibi> I think the original prbolem with the bp was the extra DB query needed for the notification
17:04:09 <gibi> and that is indeed needed
17:04:11 <gibi> see https://review.openstack.org/#/c/559076/1/nova/compute/utils.py
17:04:11 <mriedem> i see another problem
17:06:42 <mriedem> comment inline
17:06:48 <mriedem> dansmith would not like that upcall
17:07:09 <gibi> ohh, we need an upcall as well? yeah, that is a real problem
17:07:55 <mriedem> so,
17:08:21 <mriedem> 1. i don't see what need we have for this besides maybe takashi is trying to mirror the server group member ADD notification with a REMOVE notification, but there is nothing in the bp saying people are actually asking for this
17:08:38 <mriedem> 2. the bp notes mention that we don't need the remove member notification for the overquota recheck thing in the api,
17:08:51 <mriedem> 3. the upcall for the server delete case is not something we want to add, especially for something that no users are asking for
17:09:23 <gibi> I agree to ask takashi about the real use case
17:09:53 <mriedem> i could also reply to your notifications update email
17:09:56 <gibi> assuming there is one, can we just pass down the server group information from the super conductor to the compute manager during delete
17:10:01 <gibi> ?
17:10:22 <gibi> mriedem: asking takashin on the ML is OK for me
17:10:27 <mriedem> probably, but i'd like to avoid having to do that if we don't really need this
17:10:35 <gibi> mriedem: agree
17:10:57 <mriedem> i've got one thing
17:11:19 <gibi> go ahead
17:11:41 <mriedem> https://blueprints.launchpad.net/nova/+spec/add-full-traceback-to-error-notifications - i wanted to mention, if you're too busy for that, Kevin_Zheng might be interested in taking it on since his request_id in notifications blueprint is complete now
17:11:49 <mriedem> we'd have to ask him, since it's just a thought
17:12:16 <mriedem> he does have another blueprint he hasn't started coding yet so i'm not sure about his availability
17:12:34 <gibi> good point. I'm pusshing that bp down on my list of things for a while now
17:12:45 <mriedem> right, i figured with the bw based scheduling stuff....
17:12:49 <mriedem> that's going to take up all of your time
17:12:49 <gibi> I will ask Kevin_Zheng
17:12:58 <mriedem> cool
17:13:06 <gibi> yeah, the bw stuff is a heavy one
17:13:16 <mriedem> i saw the etherpad...
17:13:48 <gibi> we have a good hangout with jaypipes yesterday that cleaned up a list of things
17:14:09 <gibi> sothe next spec version will be a good step forward I think
17:14:24 <gibi> anyhow it is a lot of work
17:14:32 <mriedem> ok good, i've been waiting to read the latest update because of the recent churn
17:14:50 <mriedem> that's all i have for notifications
17:14:51 <gibi> I promise that the next version will be shorter with a bunch of outscoping ;)
17:14:57 <gibi> I have nothing else either
17:15:00 <gibi> so let's close this
17:15:04 <mriedem> o/
17:15:04 <gibi> thanks for the meeting
17:15:09 <gibi> #endmeeting