16:09:07 <sridhar_ram> #startmeeting tacker
16:09:08 <openstack> Meeting started Thu Oct 22 16:09:07 2015 UTC and is due to finish in 60 minutes.  The chair is sridhar_ram. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:09:09 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:09:11 <openstack> The meeting name has been set to 'tacker'
16:09:19 <sridhar_ram> #topic Roll Call
16:09:25 <sridhar_ram> who is here for tacker?
16:09:29 <s3wong> hello
16:09:33 <sripriya> o/
16:09:37 * sridhar_ram sorry for the delay...
16:09:49 <tbh> o/
16:10:18 <sridhar_ram> #topic Agenda
16:10:26 <sridhar_ram> #link https://wiki.openstack.org/wiki/Meetings/Tacker#Meeting_Oct_22.2C_2015
16:10:37 <sridhar_ram> anything else beyond this to discuss today ?
16:11:18 <sridhar_ram> #topic Announcements
16:11:51 <sridhar_ram> Mitaka Summit next week..most tacker devs will be away for the summit
16:12:00 <sridhar_ram> lets cancel next week's IRC meeting
16:12:13 <sripriya> +1
16:12:36 <sridhar_ram> going forward I also have an ongoing clash on this exact slot to attend TOSCA TC meeting...
16:12:45 <s3wong> sridhar_ram: really? you don't want to have us sitting in front of computer late at night in Tokyo?
16:12:57 <sridhar_ram> s3wong: :)
16:13:17 <sridhar_ram> I would like to move it to another slot...
16:13:40 <sridhar_ram> let me start a doodle poll and also discuss f2f next week
16:13:55 <s3wong> sridhar_ram: yeah, with all these new projects, finding an available timeslot and channel that works for everyone would be challenging...
16:14:16 <sridhar_ram> s3wong: true, it is going to be a nightmare, frankly..
16:14:52 <sridhar_ram> But TOSCA TC mtg is off great value for Tacker team and I don't have control to change its timing :)
16:14:59 <s3wong> sridhar_ram: last time I did it (just to move it earlier by an hour), it took a while to locate the slot/channel
16:15:31 <sridhar_ram> s3wong: need you are magic this time as well :)
16:15:39 <sridhar_ram> s/you/your/
16:16:32 <sridhar_ram> Mitaka Summit .. here is the plan for Tacker related events
16:16:35 <sridhar_ram> #link https://wiki.openstack.org/wiki/Tacker/MitakaSummit
16:17:20 <sridhar_ram> please promote it to your nfv-peeps to get the relevant audience
16:17:52 <sridhar_ram> tbh: you will be missed in next week's f2f mtg.. will try to get a mtg summary in the etherpad
16:18:09 <sridhar_ram> here at #linkhttps://etherpad.openstack.org/p/mitaka-tacker-design-summit
16:18:13 <sridhar_ram> #link https://etherpad.openstack.org/p/mitaka-tacker-design-summit
16:18:17 <tbh> sridhar_ram, yeah, thanks
16:19:05 <sridhar_ram> bobh is now officially part of tacker-core team! welcome bobh!
16:19:36 <sridhar_ram> looking forward to see what would be the icebreaker +2/+A from you :)
16:19:39 <sripriya> congratulations bobh!
16:19:55 <s3wong> or his first -2 :-)
16:20:13 <sridhar_ram> s3wong: damn.. don't plan ideas
16:20:21 <sripriya> lol
16:20:27 <tbh> congrats bobh
16:21:23 <sridhar_ram> #topic Bugs
16:21:38 <sridhar_ram> Lets zoom in high priority bugs...
16:21:41 <sridhar_ram> #link https://bugs.launchpad.net/tacker/+bugs?field.tag=liberty-critical&orderby=status&start=0
16:21:56 <sridhar_ram> we have 5 open that are marked 'liberty-critical'
16:22:28 <sridhar_ram> lets discuss few of them...
16:22:32 <sridhar_ram> https://bugs.launchpad.net/tacker/+bug/1506111
16:22:32 <openstack> Launchpad bug 1506111 in tacker "Unable to configure VDU using openwrt mgmt driver" [High,In progress] - Assigned to bharaththiruveedula (bharath-ves)
16:22:51 <sridhar_ram> tbh: can you give a quick update ? this is marked as high (which I agree)
16:23:27 <tbh> sridhar_ram, updated the patch by removing the code in create()
16:24:11 <tbh> sridhar_ram, but need the code in update() , because we need the device['attributes']['config'] in openwrt
16:24:29 <tbh> sridhar_ram, which we are setting in update() in heat.py
16:25:03 <sridhar_ram> tbh: okay.. leave that in, with your current change does the firewall config gets applied in all scenarios ?
16:25:27 <sridhar_ram> config in file & direct-input and both vnf-create & vnf-update ?
16:25:28 <tbh> sridhar_ram, yes, I have tested in all four cases
16:25:53 <sridhar_ram> tbh: lets plan to wrap this up & merge immediately after this call...
16:25:55 <sripriya> sridhar_ram: in the update case, does it need to handle the original config attribute already being specified?
16:26:29 <sripriya> can vnf-update just provide a new config for a vnf instance or should always update whatever is the current config on the vnf
16:27:07 <sridhar_ram> sripriya: I think there is a "diff" logic in the code path to see if the config has changed..
16:27:14 <sripriya> the heat update logic is basically doing a diff of that
16:27:26 <sripriya> sridhar_ram: yes
16:27:49 <sridhar_ram> sripriya: that's a reasonable logic to leave it in place.. not sure if that got validated though
16:28:11 <sripriya> sridhar_ram: ok, tbh: can you try the workflow if possible?
16:28:18 <sridhar_ram> sripriya: let tbh wrap the current set of mgmt-driver use case..
16:28:31 <sridhar_ram> we can take up other scenarios as separate effort
16:28:34 <sripriya> sridhar_ram: sure
16:28:50 <tbh> sripriya, sure
16:29:20 <sridhar_ram> we badly need functional tests in the area of mgmt_driver.. post summit we need to get few things in place
16:30:06 <sridhar_ram> tbh: sripriya: again, just to be clear .. lets not add config "diff" to the scope of the current patchset in flight...lets take it up in a separate patchset
16:30:26 <sripriya> sridhar_ram: ack
16:30:40 <sridhar_ram> tbh: make sense ?
16:30:48 <tbh> sridhar_ram, you mean not to change in update()
16:31:04 <sridhar_ram> yes...
16:31:13 <tbh> sridhar_ram, okay
16:31:33 <sridhar_ram> anything else to discuss on mgmt_driver front ?
16:32:38 <sridhar_ram> lets talk about issues in health monitor...
16:32:55 <sridhar_ram> there is an intermittent gate issue w/ health-mon
16:33:36 * sridhar_ram looking up gerrit
16:33:54 <sridhar_ram> #link https://review.openstack.org/#/c/234543/
16:34:28 <sridhar_ram> I don't think there is a bug opened yet...
16:35:34 <sridhar_ram> sripriya did some investigation... looks like a humongous tacker-heat-nova-neutron issue
16:35:52 <sridhar_ram> sripriya: did I sum it up correctly ?
16:35:55 <sridhar_ram> ;-)
16:36:58 <sripriya> sridhar_ram: yeah just to add in, related to the timing of ports addition and deletion between neutron and nova
16:38:11 <sripriya> sridhar_ram: currently doing some stress testing by adding a sleep of 3 sec between creation and deletion
16:38:26 <sripriya> sridhar_ram: that issue is not observed yet
16:38:40 <sridhar_ram> bobh: we need you are guidance in the respawn logic to see if anything can be altered in the spawn-new & delete-old sequence to avoid this
16:38:49 <sridhar_ram> sripriya: that's a good sign!
16:39:28 <sripriya> sridhar_ram: the currently logic is not altered heavily except adding a sleep
16:39:40 <sridhar_ram> sripriya: that's good...
16:39:50 <bobh> sridhar_ram: sorry I'm late, minor catastrophe here
16:40:09 <sridhar_ram> bobh: hi there, np.. hope things are fine
16:40:19 <bobh> sridhar_ram: I can take a look at respawn logic and we can discuss next week
16:40:30 <bobh> sridhar_ram: OpenStack doesn't like losing power....
16:41:02 <sripriya> sridhar_ram: not sure, why nova picks up the port deletion of old instance as a new network event on the new instance
16:41:02 <sridhar_ram> bobh: OpenStack in general doesn't like corner-cases ;-)
16:41:18 <sridhar_ram> power-outage is a corner case
16:41:27 <sripriya> even though instance_ids are different
16:41:30 <bobh> sridhar_ram: it's a big corner...
16:41:57 <sridhar_ram> bobh: yep.. your restartability patch is a good one..
16:42:20 <sridhar_ram> bobh: I remember fixing things in neutron vpnaas for neutron-server restart...
16:42:37 <bobh> sridhar_ram: catching up - not surprised there are issues with timing between nova/neutron
16:42:48 <sridhar_ram> another issue that showed up is a when both mgmt-driver and health-mon are enabled... respawn doesn't happen
16:43:22 <sridhar_ram> _mark_device_dead() fails w/ "no device" and the whole respawn logic is skipped
16:43:30 <bobh> sridhar_ram: sounds like a requirements issue :-)
16:43:37 <sridhar_ram> bobh: :)
16:44:18 <sridhar_ram> so to sum up.. we have 3 critical issues...
16:44:33 <sridhar_ram> 1) mgmt_driver changes from tbh - looks almost done
16:44:51 <sridhar_ram> 2) intermittent gate issue - promising hack from sripriya
16:45:08 <sridhar_ram> 3) mgmt_driver & respawn issue - needs further investigation
16:46:04 <sridhar_ram> I'll pull a stable/liberty EOD (Pacific) .. if we can get  (1) and (2) in it will be great
16:46:46 <bobh> sridhar_ram: I'll take a look at (3) on the plane
16:47:01 <sripriya> bobh: :-)
16:47:28 <sridhar_ram> bobh: alright.. that's sounds like a awsm plan.. sipping a glass of wine on the side I believe ?
16:48:05 <bobh> sridhar_ram: I think it's saki on ANA
16:48:22 <sridhar_ram> oh yeah, looking fwd for Sake
16:48:38 <sridhar_ram> anything else on bugs?
16:49:00 <sridhar_ram> #topic Documentation
16:49:24 <sridhar_ram> we now have tox docs jobs enabled for tacker...
16:49:36 <sridhar_ram> all the rst files are build nicely into html files
16:49:56 <sridhar_ram> See #link http://docs-draft.openstack.org/85/238385/1/check/gate-tacker-docs/d4d8065//doc/build/html/
16:50:20 <bobh> sridhar_ram: cool!
16:50:36 <sridhar_ram> this needs to be posted to readthedocs.org..
16:50:50 <sridhar_ram> there are jobs in project-config that I'm still fumbling with...
16:51:03 <sridhar_ram> sripriya: any updates from your side on docs ?
16:51:50 <sripriya> sridhar_ram: i believe that is it, i need to start working on deployment guide now that we found a place for tacker docs
16:52:55 <sridhar_ram> sripriya: thanks..
16:53:21 <sridhar_ram> I think this would be reasonable short-term solution to host tacker docs...
16:53:28 <sridhar_ram> ... in readthedocs.org
16:53:42 <sridhar_ram> unfortunately tacker.readthedocs.org is taken by another project..
16:53:49 <sridhar_ram> need to find another URL home
16:53:57 <sridhar_ram> anything else on docs ?
16:55:07 <sripriya> sridhar_ram: should we include the rest of .rst files in devref in index.rst
16:55:14 <s3wong> 5 minutes...
16:56:06 <sridhar_ram> sripriya: I've https://review.openstack.org/238385 out for index.rst and other cleanups
16:56:13 <sridhar_ram> #topic Open Discussion
16:56:17 <sridhar_ram> Xin_: ping
16:56:37 <sripriya> sridhar_ram: cool
16:56:53 <sridhar_ram> Xin_: anyone from Comcast planning to stop by Tacker BoF or developer meetup ?
16:58:08 <sridhar_ram> nothing else from my side.. for the folks traveling.. have a safe flight..see you in Tokyo!
16:58:23 <vishwana_> see you all in Tokyo soon
16:58:28 <sripriya> see you all tackers!
16:58:35 <Xin_> I do not think we will have anyone there
16:58:49 <s3wong> see (most of) you guys in Tokyo next week!
16:58:50 <Xin_> will try to make it next time
16:58:58 <sridhar_ram> Xin_: np.. we can sync up post-summit
16:59:05 <Xin_> sure..
16:59:32 <sridhar_ram> Sayonara!
16:59:40 <sridhar_ram> #endmeeting