13:00:08 <yanyanhu> #startmeeting senlin
13:00:09 <openstack> Meeting started Tue Feb 21 13:00:08 2017 UTC and is due to finish in 60 minutes.  The chair is yanyanhu. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:00:10 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:00:12 <openstack> The meeting name has been set to 'senlin'
13:00:17 <yanyanhu> hi
13:00:31 <ruijie_> hi yanyanhu
13:00:37 <yanyanhu> hi, ruijie_, evening
13:00:38 <Qiming> hello
13:00:42 <yanyanhu> hi, Qiming
13:01:01 <ruijie_> Evening all :)
13:01:07 <yanyanhu> :)
13:01:15 <yanyanhu> lets wait for a while for other attenders
13:02:32 <yanyanhu> ok, lets get started
13:02:35 <yanyanhu> https://wiki.openstack.org/wiki/Meetings/SenlinAgenda#Agenda_.282017-02-21_1300_UTC.29
13:02:38 <elynn> o/
13:02:43 <yanyanhu> hi, elynn
13:02:47 <yanyanhu> https://wiki.openstack.org/wiki/Meetings/SenlinAgenda#Agenda_.282017-02-21_1300_UTC.29
13:02:51 <yanyanhu> here is the agenda
13:02:55 <yanyanhu> please feel free to add items
13:03:09 <yanyanhu> #topic Ocata workitems
13:03:18 <yanyanhu> https://etherpad.openstack.org/p/senlin-ocata-workitems
13:03:26 <yanyanhu> the master is actually pike now
13:03:46 <yanyanhu> and ocata final release will be done in this week
13:03:59 <yanyanhu> lets go through the list
13:04:11 <yanyanhu> "Feature Rich" Nova Server
13:04:17 <yanyanhu> hi, elynn, your turn
13:04:21 <elynn> okay
13:04:34 <elynn> The patch for ports update is under reviewing
13:04:40 <elynn> I update it yesterday.
13:04:41 <yanyanhu> https://review.openstack.org/432587
13:04:42 <yanyanhu> this one
13:05:11 <elynn> It is a little complicated.
13:05:25 <yanyanhu> about?
13:05:36 <elynn> about the ports update logic
13:06:07 <elynn> But could work I think, I tested it in some simple cases.
13:06:09 <yanyanhu> so what kind of update will be supported in this initial version?
13:06:25 <yanyanhu> e.g. replacing an existing port with a new one
13:06:27 <yanyanhu> for VM
13:06:39 <elynn> For now, only replacing will supported.
13:06:46 <elynn> be supported.
13:06:47 <yanyanhu> I see
13:07:00 <yanyanhu> so the final goal is support the update of network spec
13:07:11 <yanyanhu> I mean entire network property of vm
13:07:45 <elynn> yea
13:07:58 <yanyanhu> although that seems very difficult for some cases
13:08:04 <elynn> haven't figure out how to update an existing port, since using network property to find existing port is not easy.
13:08:49 <elynn> But for vdu scenario , update is not the important part.
13:08:49 <yanyanhu> if the existing vm port was defined as "port" directly, that is easy, right?
13:09:07 <yanyanhu> the deployment is more critical?
13:09:20 <elynn> yes, yanyanhu , that's one case
13:09:32 <elynn> Yes, the deployment is more critical.
13:09:45 <yanyanhu> I see
13:09:46 <elynn> Especially the network orchestration.
13:10:15 <yanyanhu> how is the progress of deployment support?
13:10:27 <elynn> Bran has a template
13:10:54 <yanyanhu> or maybe we can temporarily put the update work aside before the deployment support can meet our requirement?
13:10:59 <elynn> to orchestrate senlin vdu cluster and also routers , networks, security groups.
13:11:28 <yanyanhu> you guys can focus on the orchestration part at the beginning I guess?
13:12:04 <elynn> Yes, I debug this template with bran these days.
13:12:17 <elynn> I think still some problems need to fix in heat...
13:12:43 <elynn> For senile profile, it's enough for now.
13:12:55 <yanyanhu> in heat? You mean the entire deployment progress will be initialized by heat
13:13:00 <yanyanhu> initialized and managed by heat
13:13:14 <elynn> Some problems about the sequence for creating and deleting resources.
13:13:14 <yanyanhu> and senlin cluster/node deployment will be part of it
13:13:25 <elynn> yanyanhu, yes
13:13:32 <yanyanhu> I see
13:13:43 <elynn> The template will create needed routers, security groups
13:13:48 <elynn> and also networks
13:13:58 <yanyanhu> yes, for routers and networks
13:14:07 <elynn> And then using them to fill in the senile profile.
13:14:15 <yanyanhu> understand
13:14:30 <yanyanhu> oh, btw, have you touched the application deployment yet?
13:14:41 <elynn> Then senlin will create ports floating IPs.
13:14:43 <yanyanhu> I mean those nfv application running inside vms
13:15:04 <elynn> They have some simple user_data, I think bran haven't touch that yet.
13:15:11 <yanyanhu> ok
13:15:23 <yanyanhu> umm, seems a lot of work to do
13:15:42 <yanyanhu> so maybe you can consider to focus on building the workflow of vdu deployment first
13:15:45 <elynn> yes, some debug work need to be done.
13:16:01 <yanyanhu> after it can be done successfully, then we start to work on update functionality
13:16:25 <yanyanhu> since network property update is really very complicated I think...
13:16:52 <elynn> yes
13:17:22 <elynn> we can have a basic one
13:17:25 <elynn> then evolve it.
13:17:29 <yanyanhu> yes
13:17:48 <yanyanhu> having a basic one is important
13:18:20 <yanyanhu> so thanks a lot for working on this and I think you can sync with xinhui to ask for her idea as well
13:18:29 <yanyanhu> she is in atlanta now I guess :)
13:18:29 <elynn> Add some todos to my patch later :)
13:18:39 <yanyanhu> elynn, thanks :)
13:18:47 <elynn> np
13:18:55 <elynn> just doing my job.
13:19:03 <yanyanhu> :)
13:19:12 <yanyanhu> ok, any more question about this item?
13:19:32 <yanyanhu> lets move on
13:19:34 <yanyanhu> Engine Improvement
13:19:43 <yanyanhu> xuefeng is not here?
13:20:09 <yanyanhu> some works have been done in last week I think
13:20:21 <yanyanhu> especially some bug fixes about health manager
13:20:23 <Qiming> yep, 431401 is in
13:20:46 <yanyanhu> yes
13:20:55 <Qiming> 424073 is in
13:21:05 <yanyanhu> but we will keep looking at/testing it
13:21:35 <yanyanhu> ok
13:21:50 <yanyanhu> and those two about health resigtry claim bug
13:22:14 <Qiming> one thing left is about the reduction of CLUSTER_CHECK/NODE_CHECK actions
13:22:36 <yanyanhu> https://review.openstack.org/435822
13:22:37 <yanyanhu> this one
13:22:53 <Qiming> right
13:23:41 <yanyanhu> hi, Qiming, should we backport this patch to ocata/stable?
13:23:42 <yanyanhu> https://review.openstack.org/435658
13:24:00 <yanyanhu> it is part of health registry fix
13:24:11 <yanyanhu> and this one https://review.openstack.org/436060
13:24:26 <Qiming> looking
13:25:46 <Qiming> em, all critical bug fixes have been backported I think
13:25:56 <yanyanhu> nice
13:26:05 <yanyanhu> so we are ready for shipping ocata release
13:26:45 <yanyanhu> and btw, many thanks for your help on backporting all those patches last friday :)
13:26:45 <Qiming> https://review.openstack.org/#/q/project:openstack/senlin+status:open
13:27:15 <yanyanhu> yes, looks pretty clean now
13:27:15 <Qiming> just a few of them needs some extra merge work
13:27:35 <Qiming> most were just a few mouse clicks and idle wait
13:27:43 <yanyanhu> great :)
13:28:19 <yanyanhu> so, lets move to next item?
13:28:34 <yanyanhu> Add senlin to RDO
13:28:34 <Qiming> np
13:28:43 <yanyanhu> no progress yet I guess
13:28:53 <yanyanhu> I think xuefeng hasn't got time to work on it yet
13:29:12 <yanyanhu> next one
13:29:13 <yanyanhu> Health Management
13:29:21 <Qiming> it'd be great if someone from from ZTE can sign on that role
13:29:36 <yanyanhu> yes, last meeting, xuefeng mentioned that
13:29:40 <Qiming> as a maintainer (focal point) of senlin in RDO
13:29:45 <yanyanhu> so hope some one can grant some hands here
13:30:00 <Qiming> it won't be a huge workload
13:30:24 <yanyanhu> just like what zigo did in debian?
13:30:29 <Qiming> the liaison will be pinged if things are broken at rdo side
13:30:39 <Qiming> we don't need packaging them at all
13:30:45 <yanyanhu> I see
13:30:49 <Qiming> generating rpm is straighforward
13:31:18 <yanyanhu> will talk with xuefeng later to see how he think about it
13:31:46 <yanyanhu> for health management, I guess there is also no progess?
13:31:56 <yanyanhu> since xinhui is traveling for ptg
13:32:03 <yanyanhu> hope she can bring us some news
13:32:15 <Qiming> no progress from my side
13:32:25 <yanyanhu> especially those work collaborated with mistral and also octavia
13:32:28 <Qiming> mistral workflow driver is in
13:32:28 <yanyanhu> Qiming, ok
13:32:45 <yanyanhu> ok
13:32:47 <Qiming> right, we need to sync with her next week
13:32:57 <yanyanhu> yes, maybe we can find a time to talk with her
13:33:03 <yanyanhu> through call or mail
13:33:34 <yanyanhu> to see the progress in her side and whether she needs any help
13:33:41 <yanyanhu> on the collaboration
13:34:04 <yanyanhu> ok, lets move on
13:34:11 <yanyanhu> Documentation
13:34:21 <yanyanhu> Qiming, I think you added lots of doc last week
13:34:25 <yanyanhu> for those policies
13:34:33 <yanyanhu> and they are all in the final rc now
13:34:46 <Qiming> em ... I think all policy types have a basic reference doc now
13:34:52 <yanyanhu> cool!
13:34:56 <Qiming> we can improve them later
13:35:01 <yanyanhu> that will be very helpful for users and also developers
13:35:04 <yanyanhu> yes
13:35:10 <yanyanhu> many thanks
13:35:20 <Qiming> I took the liberty to merge them quickly so they can be released in time
13:35:38 <yanyanhu> sure
13:35:42 <yanyanhu> no problem
13:36:20 <yanyanhu> ok, senlinclient
13:36:31 <yanyanhu> tried to setup the gate job for senlinclient functional test
13:36:32 <yanyanhu> https://review.openstack.org/#/c/436338/
13:36:38 <yanyanhu> the patch is still under review
13:36:54 <yanyanhu> after that is done, we can work on functional test cases locally
13:36:58 <Qiming> need another +2
13:37:01 <yanyanhu> and add it to tox test list
13:37:15 <yanyanhu> Qiming, yes
13:37:20 <yanyanhu> hope it works as expected
13:37:30 <yanyanhu> will talk with xuefeng about it
13:37:37 <yanyanhu> to give him some guide
13:38:04 <yanyanhu> and those two patches listed in the etherpad have been done I think
13:38:12 <yanyanhu> so will remove them from the list
13:38:41 <yanyanhu> ok, next one
13:38:45 <yanyanhu> profile for container
13:38:53 <yanyanhu> haiwei is not here
13:39:07 <yanyanhu> guess we need more attention on this work in pike cycle
13:39:20 <Qiming> yup
13:39:25 <Qiming> we need hands on it
13:40:18 <yanyanhu> lets think more about it
13:40:53 <yanyanhu> sorry, just disconnected
13:41:02 <yanyanhu> win10 is insanely unstable...
13:41:24 <Qiming> another thing we may want to do is about a refresh of the core team
13:41:25 <yanyanhu> ok, that's all items in the list I think
13:41:44 <yanyanhu> Qiming, yes, any thoughts about this
13:42:22 <Qiming> or we encourage commits/reviews first
13:42:37 <yanyanhu> yes, that sounds reasonable
13:42:42 <Qiming> there are quite a few idle cores
13:43:16 <yanyanhu> maybe we can sync with them to see their current status
13:43:22 <Qiming> let's keep thinking about this
13:43:28 <yanyanhu> ok
13:43:43 <yanyanhu> ok, lets move to next topic
13:43:52 <yanyanhu> #topic Ocata final rc and Ocata release
13:44:08 <yanyanhu> so ocata final rc has been done and final release will be done by the end of this week
13:44:17 <yanyanhu> https://releases.openstack.org/ocata/schedule.html
13:44:25 <yanyanhu> then we formally enter pike cycle I guess?
13:44:45 <Qiming> yep, starting next week I think
13:44:46 <yanyanhu> anyway, the current master is pike now
13:44:58 <yanyanhu> it will last 8 month?
13:45:03 <yanyanhu> or 6 month as usual
13:45:13 <Qiming> in theory, all work done during the the past weeks should be released in pike
13:45:13 <yanyanhu> since ocata only last 4 month I think
13:45:20 <yanyanhu> Qiming, yes
13:45:22 <Qiming> don't know
13:45:27 <Qiming> it is a chaos
13:45:33 <yanyanhu> ok...
13:45:35 <Qiming> really lost
13:45:39 <yanyanhu> lets wait for the formal schedule
13:45:44 <Qiming> don't know what the community is heading
13:45:55 <Qiming> s/what/where/
13:46:10 <yanyanhu> #topci open discussion
13:46:14 <yanyanhu> open discussion now
13:46:22 <yanyanhu> #topic open discussion
13:47:00 <yanyanhu> any more topics from you guys?
13:47:04 <Qiming> I don't have more
13:47:19 <Qiming> trapped by a lot of k8s details
13:47:38 <yanyanhu> ...
13:47:48 <yanyanhu> that is a hard work...
13:47:52 <yanyanhu> good luck for us
13:48:09 <Qiming> it is not for the poc
13:48:17 <yanyanhu> I'm also trapped by some stuff... really headache...
13:48:24 <Qiming> it is about an emerging trend
13:48:26 <yanyanhu> for accelerator support?
13:48:36 <Qiming> related
13:48:41 <yanyanhu> ok
13:48:43 <elynn> So you are working on k8s?
13:49:14 <Qiming> just started
13:49:18 <Qiming> have to move on
13:49:33 <yanyanhu> new start point :)
13:49:35 <Qiming> losing my faith gradually on openstack
13:49:36 <yanyanhu> for all of us
13:50:14 <elynn> :) have you think about any possibility to combine senlin and k8s?
13:50:27 <Qiming> looking at that as well
13:50:35 <Qiming> k8s has a cloudprovider model
13:50:41 <Qiming> where it needs a cluster api
13:50:46 <Qiming> it is empty there
13:51:04 <Qiming> the barrier is that we don't have a master node from any clusters today
13:51:11 <Qiming> that can be fixed
13:51:27 <Qiming> but for more details, you have to wait, ;P
13:51:44 <elynn> hmm, willing to learn more from you guys :)
13:51:53 <elynn> When you are ready to share.
13:52:12 <yanyanhu> :)
13:52:22 <yanyanhu> ok, I think that's all for today
13:52:29 <yanyanhu> thanks all you guys for joining
13:52:32 <Qiming> there are many fields the code quality is good, there are other areas ... not the case
13:52:40 <Qiming> good night guys
13:52:48 <elynn> good night!
13:52:49 <Qiming> will keep working on senlin anyway
13:52:51 <ruijie> have a good night :)
13:52:51 <yanyanhu> Qiming, have the same feeling
13:52:52 <Qiming> not for openstack
13:53:15 <yanyanhu> have a good night too
13:53:20 <yanyanhu> talk to you later
13:53:28 <Qiming> bye
13:53:34 <yanyanhu> bye
13:53:36 <yanyanhu> #endmeeting