13:02:06 <joehuang> #startmeeting tricircle
13:02:07 <openstack> Meeting started Wed Jan 13 13:02:06 2016 UTC and is due to finish in 60 minutes.  The chair is joehuang. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:02:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:02:11 <openstack> The meeting name has been set to 'tricircle'
13:02:44 <joehuang> hi, Yipei, welcome to join tricircle, could you pls introduce your self?
13:03:06 <Yipei_> OK. Hi everybody, My name is Yipei
13:03:40 <Yipei_> I am a master student of Huazhong university of science and technology
13:04:10 <joehuang> great, I graduated from HUST too
13:04:11 <zhiyuan> welcome to join!
13:04:42 <joehuang> #topic rollcall
13:04:51 <joehuang> #info joehuang
13:04:56 <zhiyuan> #info zhiyuan
13:05:10 <Yipei_> really?! Which year?
13:05:21 <joehuang> 1999
13:05:31 <joehuang> old enough
13:06:18 <joehuang> #topic phase2 progress
13:06:32 <joehuang> #link https://etherpad.openstack.org/p/TricircleToDo
13:06:38 <Yipei_> you must be experienced
13:06:50 <joehuang> let's look at the todolist
13:06:58 <Yipei_> OK
13:07:44 <joehuang> #info cinder volume CRUD patch was merged today
13:08:19 <joehuang> Zhiyuan, how about your patch
13:08:51 <zhiyuan> I just submitted a refined patch, the exception when operating bottom OpenStack is considered
13:09:28 <joehuang> ok, review tomorrow.
13:10:18 <zhiyuan> when neutron plugin fails to attach interfaces in bottom pod, it will try to delete the interfaces on top and bottom
13:10:54 <zhiyuan> if deleting operation fails again, the top interface will remain unbound, user can use it to attach again
13:11:59 <joehuang> any resource either could be rolled back, or could be reused again.
13:12:27 <joehuang> if something failed during the operation to bottom openstack
13:12:48 <joehuang> I think this could be one of our design principle
13:13:23 <joehuang> #info any resource either could be rolled back, or could be reused again if something failed during the operation to bottom openstack
13:14:42 <joehuang> hi Yipei, which feature do you want to be the first step for your, or some bugs?
13:15:22 <joehuang> We can help you to be family with adding code to tricircle
13:15:34 <Yipei_> I am still trying to get familiar with the project
13:15:55 <joehuang> more help needed?
13:16:05 <Yipei_> Yes
13:16:16 <joehuang> for example
13:16:27 <Yipei_> I read the blueprint of tircircle
13:16:55 <joehuang> any question
13:17:25 <Yipei_> i feel much better than solely reading source code
13:17:50 <Yipei_> but i just don't know where and how to get started
13:18:17 <joehuang> you can try playing tricircle with devstack
13:18:49 <joehuang> #link https://github.com/openstack/tricircle/tree/experiment
13:19:17 <Yipei_> OK. I will try
13:19:32 <joehuang> you can see how the process will run, receive and process request
13:19:51 <zhiyuan> to yipei, do you know the motivation of Tricircle?
13:20:17 <Yipei_> yes, sure
13:20:27 <joehuang> all services are started from the scripts located in cmd folder
13:20:57 <zhiyuan> so maybe you can start from the data model, like Pod, PodBing, ResourceRoutgin
13:21:10 <zhiyuan> s/ResourceRoutgin/ResourceRouting
13:21:34 <Yipei_> OK, got it
13:21:39 <joehuang> yes, I think Yipei can try to add some fields to the pod table,
13:22:09 <zhiyuan> then you can dig into the volume controller, I think it's easy to understand than nova, neutron, since not many resource dependency
13:22:13 <joehuang> now the "group affinity" field is missing in the pod table
13:22:45 <joehuang> yes, cinder-apigw is relative easy to understand
13:22:51 <zhiyuan> the volume controller was merged just now, as Joe said
13:24:05 <joehuang> hi, Zhiyuan, I changed the doc update to done for the task"DHCP""networking concurrency"
13:24:27 <zhiyuan> ok, thanks joe
13:25:00 <joehuang> zhiyuan, if we move the experiment to master, is it a big impact to your patch?
13:26:37 <zhiyuan> I think it's better after the l3 networking part1 patch is merged
13:26:49 <joehuang> ok
13:27:24 <zhiyuan> otherwise I am not sure whether I can figure out which files to copy to master branch
13:27:39 <joehuang> understand
13:28:01 <joehuang> our experiment has already given us very positive feedback
13:28:11 <joehuang> it's time to move experiment to master
13:28:22 <zhiyuan> that's cool
13:28:26 <joehuang> so tricircle will use stateless design in the master
13:29:03 <joehuang> #info move the experiment branch to master to adopt stateless design for Tricircle service
13:29:53 <joehuang> any other topic to be discussed?
13:30:02 <zhiyuan> oh, maybe we can start to use #openstack-tricircle channel
13:30:03 <Yipei_> next week, i plan to play tricircle with devstack and then understand the data model. if smooth, i will try to add some fields in the pod table
13:30:26 <Yipei_> how about it?
13:30:27 <joehuang> thanks Yipei
13:30:35 <zhiyuan> so yipei can discuss with joe and me online
13:30:45 <joehuang> yes good ides
13:31:05 <joehuang> #info use openstack-tricircle channel for daily communication
13:31:40 <joehuang> and I'll start the "quota" feature in phase2
13:32:16 <joehuang> because I want to find out whether there is some issue for quota control
13:32:57 <zhiyuan> ok, I will take item1 then item4
13:33:28 <joehuang> we can move VM operation to phase 3
13:34:29 <joehuang> any more topics?
13:34:38 <zhiyuan> no for me
13:34:48 <Yipei_> no for me
13:34:50 <joehuang> Yiper's suggestion is a good stat
13:35:00 <joehuang> stat -> start
13:35:02 <joehuang> ok
13:35:09 <joehuang> let's conclude the meeting
13:35:22 <joehuang> thanks for attending, have a nice day
13:35:29 <joehuang> #endmeeting