05:30:37 <sridhar_ram> #startmeeting tacker
05:30:38 <openstack> Meeting started Wed Jan 18 05:30:37 2017 UTC and is due to finish in 60 minutes.  The chair is sridhar_ram. Information about MeetBot at http://wiki.debian.org/MeetBot.
05:30:39 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
05:30:41 <openstack> The meeting name has been set to 'tacker'
05:30:45 <sridhar_ram> #topic Roll Call
05:30:49 <sridhar_ram> who is here for tacker?
05:30:52 <YanXingan> o/
05:30:53 <janki> o/
05:30:55 <diga> o/
05:30:57 <KanagarajM> hi
05:31:11 <tbh> o/
05:31:33 <gongysh> hi
05:31:41 <sridhar_ram> howdy all !!
05:31:47 <sridhar_ram> let's start..
05:31:57 <sridhar_ram> #chair gongysh tbh KanagarajM
05:31:58 <openstack> Current chairs: KanagarajM gongysh sridhar_ram tbh
05:32:16 <sridhar_ram> #topic Announcements
05:32:41 <xuhaiwei> hi
05:32:47 <sridhar_ram> #info Ocata-3 / client freeze is next week .. https://releases.openstack.org/ocata/schedule.html
05:33:11 <sridhar_ram> All client changes should merge by end of this week to make it to Ocata release
05:33:36 <sridhar_ram> please flag early if your client patchset is expected to release with Ocata
05:34:16 <sridhar_ram> next ...
05:34:41 <sridhar_ram> #info PTL self nomination is next week..
05:34:41 <sripriya> sridhar_ram: hello
05:34:53 <sridhar_ram> sripriya: xuhaiwei: hi
05:35:41 <s3wong> hello
05:35:53 <sridhar_ram> after being as your PTL for the last three cycles, I'm not planning to nominate myself.. time to pass the baton to another Tacker team member
05:36:16 <s3wong> sridhar_ram: oh...
05:36:19 <gongysh> sridhar_ram, thanks for the leading..
05:36:43 <sridhar_ram> #info sridhar_ram announces his non-candidacy for Pike
05:36:45 <sridhar_ram> gongysh: thanks
05:36:47 <s3wong> sridhar_ram: you have done well --- and Tacker really thrived under your leadership. Thanks!
05:36:47 <dkushwaha_> o/
05:37:17 <sridhar_ram> s3wong: thanks! we came a long way.. there are tons more interesting things to do
05:37:48 <sridhar_ram> I'll continue to be a core-team member..  perhaps will end up actually contributing something to the code base ;-)
05:38:07 <KanagarajM> sridhar_ram, to be frank, you are an awesome leader and brought up tacker well !
05:38:12 <gongysh> sridhar_ram, haha...
05:38:24 <KanagarajM> sridhar_ram, :)
05:38:26 <sridhar_ram> KanagarajM: thanks, means a lot
05:38:30 <tbh> sridhar_ram, great leadership till now
05:38:34 <gongysh> sridhar_ram, I have to review hard then.
05:38:46 <sridhar_ram> gongysh: damn ..
05:38:50 <sripriya> sridhar_ram: thanks for your great leadership and leading the team so effortlessly
05:38:52 * sridhar_ram running away now
05:38:59 <janki> sridhar_ram, thanks for leading...enjoyed working under your leadership :)
05:39:17 <sridhar_ram> sripriya: janki: thanks
05:39:42 <dkushwaha_> sridhar_ram: great leadership
05:39:47 <sridhar_ram> will send an email to the ML as well, thought i shd let the folks here know first
05:39:53 <sridhar_ram> dkushwaha_: thanks
05:40:11 <sridhar_ram> moving on .. i mean to the next topic ;-)
05:40:18 <sridhar_ram> #topic Ocata Status
05:40:40 <sridhar_ram> let's quickly go around outstanding Ocata work...
05:40:59 <sridhar_ram> We are officially go into the release "window" starting next week..
05:41:22 <sridhar_ram> please flag any must have bugfixes that need to make it to Ocata
05:41:31 <sridhar_ram> i'll continue to triage the LP bugs
05:41:37 <sridhar_ram> #topic NSD
05:41:59 <sridhar_ram> tbh: dkushwaha_: Part 1 merged (thanks gongysh for review / merge)
05:42:16 <sridhar_ram> tbh: dkushwaha_: can you give a status of any outstanding items ?
05:42:26 <dkushwaha_> sridhar_ram: i am just pushing code part2  with fixing your comments and some UT parts.
05:42:52 <gongysh> dkushwaha_, and some functional tests.
05:42:59 <sridhar_ram> dkushwaha_: okay
05:43:03 <tbh> sridhar_ram, and we have changes in horizon too
05:43:15 <sridhar_ram> dkushwaha_: what about this one https://review.openstack.org/#/c/334370/ ? can this be abandoned ?
05:43:27 <tbh> sridhar_ram, gongysh yes we pushed FTs but need mistral integration in gate jobs
05:43:40 <sridhar_ram> tbh: good point.. that needs some manual testing
05:43:40 <gongysh> tbh, sure
05:43:42 <dkushwaha_> sridhar_ram: gongysh I also pushed some FT in part2 . I have to work some more on UT.
05:43:50 <tbh> sridhar_ram, gongysh https://review.openstack.org/#/c/421203/
05:44:41 <sridhar_ram> tbh: cool, that should fix the gate error for NSD
05:45:24 <sridhar_ram> tbh: dkushwaha_: we also need api-ref
05:45:44 <dkushwaha_> sridhar_ram: we should add https://review.openstack.org/#/c/33470 will be abandon
05:46:03 <sridhar_ram> dkushwaha_: okay
05:46:16 <sridhar_ram> pls use the work items listed in https://blueprints.launchpad.net/tacker/+spec/nsd-support as your checklist
05:46:21 <tbh> sridhar_ram, will add api-ref for nsd
05:46:28 <sridhar_ram> tbh: thanks!
05:47:03 <sridhar_ram> we should splash this feature big in Boston .. starting thinking about writing some speaking session to use this feature !!
05:47:10 <sridhar_ram> s/starting/start/
05:47:13 <tbh> sridhar_ram, yes https://review.openstack.org/#/c/334370/ can be abandoned, will abandon this
05:47:28 <sridhar_ram> tbh: sounds good
05:47:37 <sridhar_ram> anything else on NSD?
05:48:04 <sridhar_ram> next topic..
05:48:17 <sridhar_ram> #topic VNF inline template support
05:48:18 <tbh> sridhar_ram, deadline for tacker-horizon release
05:49:16 <sridhar_ram> tbh: Feb 9th
05:49:29 <tbh> sridhar_ram, okay
05:49:58 <sridhar_ram> tacker-horizon goes along with tacker server .. Feb 9th is the deadline for both
05:50:25 <sridhar_ram> just 3.5 weeks left
05:50:31 <janki> sridhar_ram, there was discussion around the filter name on tacker channel. its fixed now. I will make changes by EOD today and submit
05:50:45 <sridhar_ram> janki: sounds good..
05:51:00 <sridhar_ram> janki: the bigger issue is the -ve cases when vnf-creation fails
05:51:13 <janki> on the server side, I will look into the error cases
05:51:36 <sridhar_ram> janki: okay.. please add some +ve and -ve func testcases as well
05:52:04 <gongysh> sridhar_ram, +ve -ve, what are they?
05:52:10 <sridhar_ram> janki: however, once you make the client changes discussed with gongysh we should get that merged
05:52:17 <janki> sridhar_ram, -ve as in?
05:52:29 <sridhar_ram> gongysh: positive and negative test cases
05:52:57 <janki> to check whether an exception is raised or not?
05:53:54 <sridhar_ram> janki: yes, when an invalid vnfd templates is passed as inline template for vnf-creation .. the API should fail appropriately (instead of the internal server error)
05:54:39 <sridhar_ram> janki: you also need to check a case where VNFD inline template is valid but the VM creation eventually fails..
05:54:58 <sridhar_ram> janki: you can use an invalid keypair name to simulate such an error
05:55:12 <KanagarajM> janki, it would be 400 , BAD REQUEST and provide additional error details in response
05:55:19 <janki> sridhar_ram, got it
05:55:56 <janki> sridhar_ram, about the vnfd-show you asked about last time, it will show for inline templtaes too. i have checked
05:56:06 <sridhar_ram> in the 2nd error scenario i described the VNF will go from PENDING_CREATE --> ERROR state
05:56:19 <sridhar_ram> janki: ack, i tried that in my testbed as well
05:57:13 <janki> sridhar_ram, cool. will add these scenarios too
05:57:23 <sridhar_ram> when a vnf-delete issued on a VNF in  ERROR state, vnfd inline template should be clean up properly
05:57:32 <sridhar_ram> that is the 2nd negative testcase
05:58:02 <sridhar_ram> anything else on this topic ?
05:58:31 <sridhar_ram> next...
05:58:43 <sridhar_ram> #topic Selective VDU scaling
05:59:07 <sridhar_ram> KanagarajM: tung_doan: spzala pushed this fix .. https://review.openstack.org/#/c/421056/
05:59:33 <sridhar_ram> KanagarajM: do you think this solves the problem that blocks selective VDU scaling ?
06:00:06 <KanagarajM> sridhar_ram, IMO, no, as my assumption is , it might fix the translation of alarm policy to Heat scaling policy
06:00:43 <sridhar_ram> KanagarajM: I see...
06:00:43 <gongysh> KanagarajM, do we have bug report for selective scaling?
06:00:46 <KanagarajM> sridhar_ram, and actual issue with HT is, failing to tx when 1> policies exist in TOSCA
06:01:23 <KanagarajM> sridhar_ram, the same bug filed earlier on HT would be ok for this as well
06:01:50 <KanagarajM> https://bugs.launchpad.net/heat-translator/+bug/1652928
06:01:50 <openstack> Launchpad bug 1652928 in Heat Translator "Fails to translate the derived polices" [High,New] - Assigned to Sahdev Zala (spzala)
06:02:22 <KanagarajM> sridhar_ram, sahdev alreay using this bug for the same.
06:03:18 <sridhar_ram> KanagarajM: is sahdev waiting for the sample HOT template from you ?
06:03:54 <KanagarajM> sridhar_ram, already given the sample
06:04:05 <sridhar_ram> KanagarajM: okay..
06:04:23 <sridhar_ram> given sahdev is going to be offline.. i'm not sure if this will be fixed for Ocata :(
06:04:41 <sridhar_ram> *offline next week
06:05:13 <sridhar_ram> bobh: tbh: do you have any visibility into this HT fix ?
06:05:13 <KanagarajM> sridhar_ram, I too think the same :(
06:06:07 <sridhar_ram> KanagarajM: assuming HT fix goes in, is there any change required in tacker to make use of it ?
06:06:08 <XueFeng> quit
06:06:14 <KanagarajM> sridhar_ram, if bobh and tbh have spare some time, it would help considering the situation.
06:06:59 <KanagarajM> sridhar_ram, yes, after HT provides the feature, we should update the openstack driver to accomedate HT new feature and strip out the corresponding code from tacker
06:06:59 <sridhar_ram> tbh has NSD in his plate
06:07:30 <tbh> sridhar_ram, KanagarajM haven't discussed with sahdev on this yet
06:07:31 <sridhar_ram> KanagarajM: okay, that good to know.. we have time till Feb 9th to make that happen
06:08:10 <sridhar_ram> next.. moving to a related topic
06:08:12 <KanagarajM> sridhar_ram, ok
06:08:19 <sridhar_ram> #topic Tacker-Senlin integration
06:08:30 <sridhar_ram> the spec is close to merge, IMO
06:08:54 <xuhaiwei> hi sridhar_ram, I have updated the spec due to the comments from cores
06:09:01 <sridhar_ram> xuhaiwei: KanagarajM: looks the only outstanding item is the support for this selective VDU scaling using Senlin
06:09:17 <xuhaiwei> I think it should have made things clear
06:09:46 <KanagarajM> sridhar_ram, yes and it was claimed that senlin does not support at this moment. I think
06:09:57 <xuhaiwei> sridhar_ram, yes, for this feature, I have confirmed with senlin team, it is under implementation
06:09:59 <sridhar_ram> xuhaiwei: yes, that clarifies
06:10:31 <sridhar_ram> KanagarajM: are you okay to proceed given this clarification ?
06:10:35 <KanagarajM> sridhar_ram, considering that we could go ahead with spec and  add corresponding depency of senlin bp
06:10:47 <xuhaiwei> KanagarajM, it is on the way, will be there in the near future
06:10:54 <sridhar_ram> KanagarajM: sounds good..
06:11:06 <KanagarajM> sridhar_ram, :) yes
06:11:35 <xuhaiwei> sridhar_ram, KanagarajM, that sounds good to me
06:11:36 <sridhar_ram> xuhaiwei: now, how far is the implementation progressed ?
06:11:56 <sridhar_ram> xuhaiwei: we just have few meeting to finish the reviews, assuming the implementation is far along
06:12:02 <xuhaiwei> sridhar_ram, you mean selective vdu scaling?
06:12:38 <sridhar_ram> xuhaiwei: no, not for selective vdu scaling.. the overal tacker-senlin autoscaling
06:12:42 <sridhar_ram> *overall
06:13:09 <xuhaiwei> sridhar_ram: I am afraid it is difficult to finish it in ocata cycle
06:13:14 <sridhar_ram> earlier i meant 'we just have few weeks to finish the reviews'
06:13:32 <sridhar_ram> xuhaiwei: I see. No worries.
06:14:03 <sridhar_ram> xuhaiwei: do continue to work on the implementation.. this could be a nice early feature to land in Pike
06:14:07 <xuhaiwei> sridhar_ram: feature freeze will be the next weekend, wont it?
06:14:10 <gongysh> sridhar_ram, we can go without selective scaling for the first draft.
06:14:33 <xuhaiwei> sridhar_ram: I will try my best to do it
06:14:36 <sridhar_ram> xuhaiwei: tacker server's deadline is Feb 9th
06:14:50 <sridhar_ram> next week deadline is for tackerclient..
06:15:03 <sridhar_ram> i don't think this senlin integration as any client side changes
06:15:07 <xuhaiwei> sridhar_ram: I see
06:15:07 <sridhar_ram> correct?
06:15:28 <sridhar_ram> xuhaiwei: you've about 3 weeks
06:15:34 <sridhar_ram> gongysh: agree
06:15:39 <KanagarajM> sridhar_ram, yes, it deoes not have any api change
06:15:41 <xuhaiwei> sridhar_ram: should be no changes in tackerclient
06:15:58 <sridhar_ram> KanagarajM: xuhaiwei: sounds good
06:16:14 * Qiming wondering whether PTG is a good thing that leads to such a short cycle for all projects
06:16:23 <sridhar_ram> let's shot this as a best effort for Ocata..
06:16:33 <sridhar_ram> *shoot
06:16:44 <xuhaiwei> sridhar_ram: ok
06:17:31 <sridhar_ram> Qiming Ocata is a one off short cycle.. it shd get better Pike and onwards
06:17:53 <Qiming> it is buying us a longer cycle for Pike? Good!
06:18:30 <sridhar_ram> Qiming: yep, it gradually springs back to 6months cycle... with PTG coinciding with a mid-cycle
06:18:41 <sridhar_ram> anything else on senlin integration ?
06:18:52 <xuhaiwei> nope from me
06:19:03 <sridhar_ram> next ..
06:19:07 <gongysh> xuhaiwei, get a draft implemenation and have a demo
06:19:07 <sridhar_ram> #topic Ocata Bugs
06:19:08 <Qiming> just one note, senlin plans to extend its nova server profile
06:19:14 <gongysh> xuhaiwei, this will help.
06:19:28 <xuhaiwei> gongysh: thanks
06:19:44 <Qiming> so that a so called "nova server" can have ports, security groups, volumes auto-created and managed if needed
06:19:45 <sridhar_ram> folks - i know we spend a lot of time on new features.. but over next few weeks I'd like to request all of you to focus on bug fixes for Ocata.
06:21:15 <sridhar_ram> please wrap the bugs  marked for Ocata 3  - https://launchpad.net/tacker/+milestone/ocata-3
06:21:41 <sridhar_ram> please help to mark critical / useful bugs with ocata-3 milestone target so that it gets enough attention
06:21:48 <sridhar_ram> thanks for you support
06:22:01 <sridhar_ram> #topic Open Discussion
06:22:10 <sridhar_ram> any thing else ?
06:23:02 <sridhar_ram> YanXingan: thanks for your participation so far.. please continue to pick bigger items like RFEs and BPs
06:23:16 <YanXingan> ok
06:23:30 <YanXingan> i will speed more time
06:23:39 <sridhar_ram> YanXingan: great
06:24:03 <janki> sridhar_ram, about the contanired bp, I will update it by end of this week. I have a draft template. will add it soon
06:24:53 <sridhar_ram> janki: sure, there are some interesting comments in that spec..
06:25:18 <sridhar_ram> Folks - please also review diga's API framework spec .. https://review.openstack.org/#/c/368511/
06:25:23 <diga> sridhar_ram: WE want to merge the API spec as soon as possible, need to start dev work, Just reminding you again
06:25:29 <diga> Yes
06:25:32 <diga> :)
06:25:38 <janki> sridhar_ram, yes. I got a guy from kuryr to review it. I will have also chat woth madhuri to review it after I update
06:25:45 <janki> *with
06:25:55 <sridhar_ram> janki: sounds good
06:26:22 <janki> :)
06:26:23 <sridhar_ram> diga: i still have some questions on the http-server you are proposing.
06:26:31 <diga> sridhar_ram: okay
06:26:54 <sridhar_ram> diga: instead of gunicorn you propose the keep the current wsgi.py in tacker as the http server ?
06:27:03 <diga> sridhar_ram: Yes
06:27:10 <sridhar_ram> gongysh: is the expert here on wsgi :)
06:27:40 <diga> gongysh: how are we using wsgi for tacker web ?
06:27:43 <sridhar_ram> diga: what is zaqar using for http-server ?
06:27:50 <diga> sridhar_ram: yes
06:28:05 <diga> sridhar_ram: wsgi
06:28:15 <gongysh> sridhar_ram,  diga, we need to be installed behind nginx or apach.
06:28:53 <diga> sridhar_ram: that's where I m coming to, we should use to nginx for tacker web
06:29:03 <diga> gongysh: yep
06:29:32 <sridhar_ram> diga: let's take it up in gerrit
06:29:35 <sridhar_ram> times up
06:29:40 <diga> sridhar_ram: okay
06:29:59 <sridhar_ram> bye everyone!
06:30:04 <janki> bye
06:30:05 <sridhar_ram> #endmeeting