15:00:22 #startmeeting openstack search 15:00:23 Meeting started Thu Jun 1 15:00:22 2017 UTC and is due to finish in 60 minutes. The chair is sjmc7. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:24 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:26 very nice 15:00:27 The meeting name has been set to 'openstack_search' 15:01:04 o/ 15:01:23 morning 15:01:36 it is morning here 15:02:00 here too but only for 1 more hour 15:02:42 i am under a coffee emergency, ttyl 15:02:51 bye! 15:03:45 I'm going to have one more tea 15:04:15 i don’t have anything specific for the agenda again, except milestone 2 is approaching next week 15:05:32 so we should be moving to bug finding and fixing mode 15:05:57 lei-zh: i promise i will look at the WSGI patch today, apologies for not doign so earlier 15:07:15 oh, TravT - one for you maybe. the UI branch is failing NPM tests (gate-searchlight-ui-nodejs4-npm-run-test) 15:07:16 sjmc7: I planed to test some patches today in devstack, but the setuptools new release breaks the whole world : ( 15:07:29 oh, fun 15:07:34 oo, i missed that. that’s always good times 15:07:48 setuptools and pbr are the lynchpins of everything 15:07:55 plus there's a problem with ansible missing on a bunch of nodes 15:08:07 so 2 bad gate problems simultaneously 15:08:19 better now than m-3 15:08:32 TravT: error is http://logs.openstack.org/12/457212/2/check/gate-searchlight-ui-nodejs4-npm-run-test/ccc7de0/console.html.gz#_2017-05-25_10_31_28_703852 15:08:48 it looks like a dependency problem but i don’t understand how npm versioning works in openstack 15:14:06 there aren’t that many open patches (at least, not current ones). the WSGI one, versioned notifications (to which i need to add a release note) 15:17:10 i added https://review.openstack.org/#/c/467641/ to the governance repo to indicate where we were with pike goals 15:18:21 yeah, I forget to add that, thanks 15:19:21 lei-zh: has there been any more forward work on using the notification forwarding? 15:20:50 TravT: not yet, still blocked by a zaqar bp 15:21:26 but I met a zaqar core working on the spec two weeks ago at a bugsmash event, have described our usecase to him 15:23:55 he said it should be finished in pike release, but I'm not sure if it's guaranteed 15:24:02 it’s openstack! :) 15:25:39 there maybe one way to work out a demo by using a zaqar presigned queue 15:28:48 i don't understand zaqar well enough to know what that means... 15:30:25 lol, it means we create a special zaqar queue, and hand a signature to consumers 15:30:58 they will use that signature to get the content of that queue, completely bypassing keystone 15:31:27 it just works for demo purpose I think 15:33:24 ship it! :) 15:34:24 I don't know if there are other directions we can have a look at, doing message queue' job is definitely not searchlight's goal 15:41:25 right, zaqar would hopefully solve the awkward tenancy 15:41:40 that said, i did once set something up that hooked into oslo messaging and maintained websocket connections by tenant 15:43:22 yeah, but it's open source : ) 15:45:51 it seems there are not so many use cases to deliver message from internal openstack service to tenants 15:47:50 the horizon one was a big one that we’d wanted for a long time 15:52:11 so what about we maintain a web socket connection directly with consumers? 15:52:39 we could do that, but as you say, it’s not really what searchlight set out to do, and it would have to be through the API process which doesn’t directly communicate with listeners 15:55:12 ok, guess i’ll end it here. let’s try to get these patches merged and a bit of testing to coax out any bugs 15:55:13 thanks all 15:55:32 #endmeeting