13:00:21 <ruijie> #startmeeting senlin
13:00:22 <openstack> Meeting started Tue Oct 17 13:00:21 2017 UTC and is due to finish in 60 minutes.  The chair is ruijie. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:00:23 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:00:25 <openstack> The meeting name has been set to 'senlin'
13:00:48 <ruijie> hi
13:01:13 <Qiming> hi, ruijie
13:01:19 <ruijie> hi Qiming
13:01:24 <ruijie> https://wiki.openstack.org/wiki/Meetings/SenlinAgenda#Agenda_.282017-10-17_1300_UTC.29
13:01:39 <ruijie> here is the agenda
13:03:00 <Qiming> okay
13:04:40 <ruijie> the work item I am following on are: https://review.openstack.org/#/c/504870/ and https://review.openstack.org/#/c/506550/
13:05:32 <ruijie> they have migrated the zuul to zuul V3, so we may need another patches to add the job templates for senlin-tempest-plugin project
13:05:33 <Qiming> looks like both are on track
13:06:03 <Qiming> okay, I believe we may need to adjust the CI scripts for gate jobs to run correctly
13:06:57 <ruijie> yes Qiming
13:07:13 <ruijie> another question is about the dead lock ..
13:07:41 <ruijie> it came up again, and as I talked with chenyb4, they met the problem too
13:09:04 <Qiming> is there a bug report on this?
13:10:03 <ruijie> not yet Qiming, I am still looking at it, looks like still the problem of DB side
13:10:14 <Qiming> sigh
13:10:29 <ruijie> the cluster lock was not released correctly
13:10:44 <Qiming> not released?
13:12:25 <ruijie> yes Qiming, triggered cluster_delete, but the action failed. The current situation is: 3 of the node_delete action are still in "RUNNING" state, the cluster action is in "FAILED" state, and the cluster lock is still there
13:13:39 <Qiming> node deletion running
13:13:45 <Qiming> but cluster action failed?
13:13:54 <ruijie> yes .. strange
13:14:10 <Qiming> that could not happen
13:14:34 <Qiming> it sounds not like a cluster lock problem
13:15:17 <Qiming> the question is why cluster action failed?
13:16:00 <ruijie> the profile layer raised exceptions, some of the node deletion actions failed
13:16:14 <Qiming> then node delete actions should fail
13:16:52 <ruijie> yes Qiming, none of the node actions shoule be in "RUNNING" state ..
13:17:36 <Qiming> so ... a bug report with detailed status of locks would be helpful
13:18:22 <ruijie> sure Qiming, I will draft an bug with the details
13:18:44 <Qiming> do we need some tooling support for better debugging?
13:19:32 <ruijie> Qiming you have any suggestions, I am using log and pdb now ..:(
13:19:40 <Qiming> the sqlalchemy layer is very annoying ...
13:20:12 <Qiming> and oslo.db layer is masking some of the parameters
13:22:19 <ruijie> maybe we can find some methods to reduce the dependency on DB?
13:22:40 <Qiming> I don't think so
13:23:08 <Qiming> DB concurrency is a mature technology
13:23:26 <Qiming> however, sqlalchemy and oslo.db is making things a lot difficult to diagnose
13:24:13 <Qiming> if we manage locks in python ... that is gonna be a reinvention of the wheel ... even if we managed to get it done right
13:26:28 <ruijie> okay Qiming, that makes sense
13:27:07 <Qiming> so .. back to the problem mentioned, I think we need to solve the problem in several steps
13:33:22 <ruijie> sir ?
13:33:32 <Qiming> ?
13:33:56 <ruijie> I thought you will talk about the stops
13:33:59 <ruijie> :)
13:34:08 <ruijie> steps
13:35:29 <ruijie> let's moving on?
13:36:30 <Qiming> sure
13:37:31 <ruijie> about the meetup, I talked with people, liyi and elynn are okay for 28th Oct. and xuefeng will not be available at that time
13:38:35 <ruijie> I didn't cache xinhui yet :) will send her another e-mail to check the schedule
13:38:43 <Qiming> great
13:39:31 <ruijie> maybe we can share the window in the meeting, so that people who cannot make it can still join the online meeting
13:40:06 <Qiming> good idea
13:40:40 <Qiming> pls let me know if there are things I can help
13:41:57 <ruijie> sure Qiming :) will update the link on etherpad https://etherpad.openstack.org/p/senlin-queens-meetup
13:43:07 <ruijie> next topic is yours Qiming
13:43:18 <Qiming> what?
13:43:29 <Qiming> summit prep?
13:43:32 <ruijie> summit prep
13:43:46 <Qiming> my trip request was rejected
13:43:46 <ruijie> yes Qiming .. we have a light talk?
13:43:53 <ruijie> me too :)
13:44:22 <Qiming> I applied for trip sponor program but failed
13:44:33 <ruijie> hahha
13:44:40 <ruijie> me either :(
13:44:53 <ruijie> not sure what happend this summit
13:45:16 <Qiming> trip to australia is expensive
13:45:56 <Qiming> some money has been allocated to PTG by the foundation I guess
13:48:18 <Qiming> elynn pinged me today, about his talk
13:48:50 <Qiming> his installation of k8s was smooth
13:49:10 <ruijie> with senlin?
13:49:23 <Qiming> although he had to regenerate the server cert so that the floating IP of the master node can be used
13:49:31 <Qiming> yes
13:49:47 <ruijie> cool
13:51:36 <Qiming> other than that, I am not aware of other topics acception
13:52:36 <ruijie> okay Qiming, we may end the meeting a little bit early
13:52:47 <Qiming> okay
13:52:49 <Qiming> thanks
13:52:59 <ruijie> thanks for joining Qiming
13:53:07 <ruijie> Goog night
13:53:14 <Qiming> u2
13:53:15 <ruijie> #endmeeting