18:00:04 <SumitNaiksatam> #startmeeting networking_policy
18:00:05 <openstack> Meeting started Thu Jul 23 18:00:04 2015 UTC and is due to finish in 60 minutes.  The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:06 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:00:09 <openstack> The meeting name has been set to 'networking_policy'
18:00:22 <SumitNaiksatam> #info agenda https://wiki.openstack.org/wiki/Meetings/GroupBasedPolicy#July_23rd_2015
18:00:37 <SumitNaiksatam> Yi: hi
18:00:47 <SumitNaiksatam> update on the kilo release - its still in the works
18:00:50 <Yi> Hi
18:01:05 <SumitNaiksatam> we are making good progress with the outstanding items
18:01:10 <SumitNaiksatam> ivar-lazzaro: hi
18:01:23 <ivar-lazzaro> hi!
18:01:30 <SumitNaiksatam> but it might take a little longer to fix some of the outstanding issues
18:01:49 <SumitNaiksatam> so we will evaluate early next week as to what date we want to do the release
18:01:58 <SumitNaiksatam> any questions/comments?
18:02:26 <mageshgv> SumitNaiksatam: Approximately when is the release planned as of now
18:02:41 <SumitNaiksatam> mageshgv: at this point, its at least a week out
18:02:51 <mageshgv> SumitNaiksatam: ok
18:03:13 <SumitNaiksatam> mageshgv: if there are any concerns with the time frame, please let me know
18:03:40 <SumitNaiksatam> and/or we can discuss here
18:03:56 <mageshgv> SumitNaiksatam: No concerns as such, but we might have to revisit the bugs and prioritise any important ones
18:04:20 <SumitNaiksatam> mageshgv: bugs are prioritized to a large extent
18:04:31 <SumitNaiksatam> mageshgv: some of those we need to fix before the release
18:04:37 <SumitNaiksatam> others we can backport
18:05:02 <mageshgv> SumitNaiksatam: Okay
18:05:09 <SumitNaiksatam> mageshgv: can have an offline discussion if you are at a point where you want to start knocking off the bugs from the list
18:05:27 <mageshgv> SumitNaiksatam: sounds good
18:05:33 <SumitNaiksatam> mageshgv: thanks
18:05:44 <SumitNaiksatam> talkkng about bugs
18:05:46 <SumitNaiksatam> #topic Bugs
18:06:05 <SumitNaiksatam> no criticals on the server side, but there is one which is on the UI side
18:06:08 <SumitNaiksatam> in stable/juno
18:06:32 <SumitNaiksatam> #link https://bugs.launchpad.net/group-based-policy-ui/+bug/1476866
18:06:33 <openstack> Launchpad bug 1476866 in Group Based Policy UI "Create member is broken in stable/juno" [Critical,Confirmed] - Assigned to ank (ank.b)
18:06:58 <SumitNaiksatam> fix has been posted but we are having problems with the py27 gate: #link https://review.openstack.org/204507
18:07:44 <SumitNaiksatam> other that that, are there any importants bugs that anyone wants to discuss in the meeting here?
18:07:48 <mageshgv> SumitNaiksatam: There is one more critical issue that I found on UI yesterday
18:07:50 <SumitNaiksatam> *important
18:08:03 <mageshgv> #link https://bugs.launchpad.net/group-based-policy-ui/+bug/1477064
18:08:04 <openstack> Launchpad bug 1477064 in Group Based Policy UI "Updating any resource does not work for non-admin user" [Undecided,New]
18:08:14 <SumitNaiksatam> mageshgv: yes i saw that
18:09:07 <mageshgv> SumitNaiksatam: This blocks us from using any of the update operations from UI today
18:09:20 <SumitNaiksatam> mageshgv: we actually had a bug for that earlier as well
18:09:25 <SumitNaiksatam> trying to find the link
18:09:50 <SumitNaiksatam> mageshgv: but agree we have to fix this asap - the idea would be to update only the attributes that are changed in teh form
18:10:15 <SumitNaiksatam> #link https://bugs.launchpad.net/group-based-policy-ui/+bug/1465345
18:10:16 <openstack> Launchpad bug 1465345 in Group Based Policy UI "Edit PTG form has limited usability" [High,Confirmed]
18:10:47 <SumitNaiksatam> mageshgv: ^^^ but i agree that this is an issue across, so i will dup the above to your bug
18:11:05 <SumitNaiksatam> mageshgv: are you or ank planning to work on this issue?
18:11:06 <mageshgv> SumitNaiksatam: Ah, I didnt notice this.
18:11:57 <mageshgv> Ankaiah may work on this, but at this moment it is not clear, may be ransari can update on that
18:12:08 <SumitNaiksatam> mageshgv: ok, i will check off line
18:12:29 <SumitNaiksatam> ank is actually working on a couple of other bugs too
18:12:40 <SumitNaiksatam> so i think he might be able to get to this after those
18:12:49 <ransari> mageshv: we will have to close off internally on this
18:12:50 <SumitNaiksatam> any other bugs that we want to discuss?
18:13:13 <mageshgv> ransari: okay
18:13:14 <ransari> U support for external policy needs to be priortized as well
18:13:24 <ransari> I don't hae the link right now
18:13:35 <SumitNaiksatam> ransari: yeah it doesnt exist
18:13:50 <ransari> clarification: UI support for external policy
18:13:54 <SumitNaiksatam> ransari: that and service profile are other critical bugs on the UI
18:14:29 <SumitNaiksatam> but currently no one signed up for fixing those
18:15:07 <SumitNaiksatam> any other bugs?
18:15:07 <ransari> SumitNaiksatam: will update you w.r.t sign up for those two
18:15:15 <SumitNaiksatam> ransari: great, thanks!
18:15:52 <SumitNaiksatam> #topic Testing
18:16:51 <SumitNaiksatam> I was able to do a run of the Rally tests in the gate from this repo: #link https://github.com/group-policy/rally/tree/dev
18:17:04 <SumitNaiksatam> see the result: #link http://logs.openstack.org/63/202263/9/check/gate-group-based-policy-dsvm-functional/0d61896/console.html
18:17:21 <SumitNaiksatam> i tried concurrency of 10 for creating every resource
18:17:32 <SumitNaiksatam> currently no failures
18:18:22 <SumitNaiksatam> you will have to scroll somewhere in the middle of the log to see the rally results
18:18:32 <rkukura> SumitNaiksatam: Any clue what to search for?
18:18:51 <SumitNaiksatam> rkukura: yeah - “Response Times (sec)"
18:19:48 <rkukura> Are these all with 10 concurrent requests?
18:19:50 <SumitNaiksatam> so at this point i am inclined to submit an infra patch to add an addtional job to run these tests on every patch
18:20:31 <SumitNaiksatam> rkukura: that is my understanding, but i am in the process of checking with Ajay that the concurrency parameter did actually take effect as expected
18:20:53 <SumitNaiksatam> there are two parameters, concurrency and count - the latter being the number of times you want to run
18:21:05 <rkukura> ok, thanks
18:21:12 <ivar-lazzaro> SumitNaiksatam: Can we have those tests in tree?
18:21:34 <ivar-lazzaro> or it would need more work?
18:21:54 <SumitNaiksatam> ivar-lazzaro: there are some changes that Ajay had made to the rally code base to be able to run these tests
18:22:21 <SumitNaiksatam> ivar-lazzaro: so there are changes in that repo beyond just adding the tests that are required to run these tests
18:22:34 <ivar-lazzaro> ok
18:22:42 <SumitNaiksatam> ivar-lazzaro: so the long term goal is to upstream those changes to rally itself
18:22:55 <SumitNaiksatam> ivar-lazzaro: but until then we can still run these tests in teh current way
18:23:19 <SumitNaiksatam> ivar-lazzaro: we could copy over just the tests to our tree, but we would still need the other repo to run them
18:23:52 <SumitNaiksatam> ivar-lazzaro: but we can check with Ajay if he is comfortable checking in the just the tests into the gbpservice tree
18:24:04 <ivar-lazzaro> I'm giving a look at the tests
18:24:09 <ivar-lazzaro> and the framework seems really nice
18:24:27 <SumitNaiksatam> i was also tempted to add these tests just as a part of the same integration job, since it already installs devstack
18:24:34 <ivar-lazzaro> shouldn't be hard to write scenario tests when you commit a patch
18:25:00 <ivar-lazzaro> if we had this in tree I feel that we could add many more tests over time
18:25:14 <SumitNaiksatam> ivar-lazzaro: by looking at how these tests are written, we can replicate for new features/additions
18:25:43 <SumitNaiksatam> ivar-lazzaro: yes, i agree, that was is indeed the plan or record, this is an intermediate step to get to that point
18:25:52 <SumitNaiksatam> * plan of
18:26:25 <SumitNaiksatam> so let me know if anyone has thoughts around making this part of the same integration job or running a completely new job for these
18:26:26 <ivar-lazzaro> Not sure I understand the success criteria though
18:26:58 <SumitNaiksatam> ivar-lazzaro: it will tell you what percentage of the iterations failed
18:27:27 <ivar-lazzaro> SumitNaiksatam: what is a failure? an exception? a result different from 20X from the APIs?
18:28:04 <SumitNaiksatam> ivar-lazzaro: yes, an error or exception condition
18:28:40 <SumitNaiksatam> so initially this test job would be non-voting if we create a separate job
18:28:41 <ivar-lazzaro> So the concurrency problem is not really investigated with this (eg. more default L3Ps created as a result of 2 EPG created)
18:29:25 <SumitNaiksatam> ivar-lazzaro: it depends on how you write the test
18:30:12 <SumitNaiksatam> creation of some resources will result in creation of other resources too, and the whole thing should happen in parallel if concurrency is configured
18:30:13 <ivar-lazzaro> I see, I've never worked with rally but this seems very good! Thanks Ajay!
18:31:10 <SumitNaiksatam> the other related item here - making the current integration job to be a voting job
18:31:35 <SumitNaiksatam> we have been treating it as criteria for approving a patch
18:31:41 <SumitNaiksatam> and it has been stable for a while now
18:31:53 <SumitNaiksatam> so i am proposing to make it voting
18:31:59 <SumitNaiksatam> let me know if there are any concerns
18:32:35 <SumitNaiksatam> okay so i will post the patch
18:32:48 <SumitNaiksatam> #topic Packaging
18:33:01 <SumitNaiksatam> rkukura: anything to discuss or update?
18:33:09 <rkukura> nothing new
18:33:18 <SumitNaiksatam> rkukura: okay
18:33:36 <SumitNaiksatam> ransari: any new requirements, or anything you want to reiterate from before?
18:35:26 <SumitNaiksatam> ok moving on
18:35:33 <SumitNaiksatam> #topic Docs
18:36:15 <SumitNaiksatam> I am in the process of posting a patch which adds a little more structure to the in-tree documentation and also tries to make it consistent with the other projects
18:36:36 <SumitNaiksatam> i am also hoping to have an API document in there
18:37:13 <SumitNaiksatam> once we have that, we can also madate updating the API or usability changes in the in-tree documentation along with the implementation patch that makes these changes
18:37:27 <SumitNaiksatam> that way we can hopefully always keep the documentation consistent
18:38:05 <SumitNaiksatam> let me know if you have any ideas on this or if you disagree
18:38:34 <SumitNaiksatam> #topic Client update
18:38:51 <SumitNaiksatam> we discussed last week about creating a stable branch for the client
18:38:59 <SumitNaiksatam> so stable/juno has been created
18:39:20 <SumitNaiksatam> and we can start backporting any relevant commits
18:39:58 <SumitNaiksatam> #topic Kilo Items
18:40:29 <SumitNaiksatam> #link https://review.openstack.org/#/c/203226 (Plumbing Terminology)
18:40:41 <SumitNaiksatam> please review the above
18:40:58 <SumitNaiksatam> ivar-lazzaro: anything you would like to bring up for discussion on this?
18:41:10 <SumitNaiksatam> or anyone else who has already review this
18:41:17 <ivar-lazzaro> I see there are new comments
18:41:24 <ivar-lazzaro> but I haven't read them yet
18:41:34 <SumitNaiksatam> is songole here?
18:41:46 <songole> Yes SumitNaiksatam
18:42:52 <SumitNaiksatam> anything you want to discuss regarding: #link https://review.openstack.org/#/c/203226, you had review comments
18:42:58 <songole> Nothing at this time.
18:43:22 <songole> I will catch up with ivar later today
18:43:51 <SumitNaiksatam> songole: okay
18:44:08 <ivar-lazzaro> yes please. The thing I'm looking at the most is new use cases or gotchas in the terminology
18:44:57 <ivar-lazzaro> we need to label services with similar plumbing requirements as much accurately as possible
18:45:04 <SumitNaiksatam> the following other patches have also been in review queue for some time - #link https://review.openstack.org/#/c/189182, #link https://review.openstack.org/166424, #link https://review.openstack.org/179327
18:45:30 <SumitNaiksatam> so please help review
18:45:42 <SumitNaiksatam> ivar-lazzaro: anything you wanted to bring up for discussion regarding the above?
18:46:24 <ivar-lazzaro> there were concerns around #link https://review.openstack.org/#/c/189182
18:46:33 <ivar-lazzaro> in the workflow especially
18:47:00 <ivar-lazzaro> But we never got to a conclusion around how the service management PTG should be defined
18:47:06 <SumitNaiksatam> ivar-lazzaro: okay
18:47:34 <SumitNaiksatam> ivar-lazzaro: you mean how its created?
18:47:37 <ivar-lazzaro> as far as #link https://review.openstack.org/#/c/166424/ is concerned
18:47:53 <ivar-lazzaro> we shouldn't merge it until rkukura driver is tested with that as a dependency
18:48:00 <ivar-lazzaro> so I'll put the WIP
18:48:24 <ivar-lazzaro> SumitNaiksatam: That, and also how is scheduled by the NCP
18:49:02 <ivar-lazzaro> I think some were suggesting to remove the new attribute, and have it explicitly set somehow during the chain creation
18:49:46 <SumitNaiksatam> ivar-lazzaro: okay, i am not sure i understand the “scheduled” terminology but i can go and check back in the review
18:49:51 <ivar-lazzaro> about https://review.openstack.org/#/c/179327/ I think rukshana tested the patch against the Neutron fix but it's still broken
18:50:20 <ivar-lazzaro> SumitNaiksatam: I mean which service management PTG you use when the chain is created
18:50:33 <SumitNaiksatam> ivar-lazzaro: okay, got it
18:51:12 <SumitNaiksatam> ivar-lazzaro: yes, i recollect ransari going through that exercise a couple of weeks back
18:51:24 <SumitNaiksatam> ivar-lazzaro: i dont think she revisited that
18:52:02 <SumitNaiksatam> ivar-lazzaro: thanks for the update and the summary
18:52:11 <SumitNaiksatam> rkukura: any update you want to share on the nova driver?
18:52:48 <rkukura> I’ve worked through the lockup launching a nova VM, but still not sure I understand why that was happening
18:52:56 <SumitNaiksatam> rkukura: okay
18:53:23 <rkukura> Seems I need to both enable multiple api_workers and do the nova call from a separate thread. Not sure why I’d need both.
18:53:59 <rkukura> At least I can make forward progress finally.
18:54:04 <SumitNaiksatam> rkukura: :-)
18:54:05 <rkukura> That’s it for now
18:54:25 <SumitNaiksatam> rkukura: thanks, yeah that sounds wierd
18:54:59 <SumitNaiksatam> i should have a patch on the Quota support for GBP resources soon
18:55:40 <SumitNaiksatam> any other feature-related items we missed out?
18:56:05 <SumitNaiksatam> igordcard_: i noticed you dramatically reduced the size of your patch? ;-)
18:56:47 <igordcard_> SumitNaiksatam: yes, it was initially based on the previous TS API, which does not apply for GBP
18:57:03 <SumitNaiksatam> igordcard_: okay, hoping to see the new version soon
18:57:07 <SumitNaiksatam> #topic Open Discussion
18:57:07 <igordcard_> SumitNaiksatam: starting from scratch regarding neutron/gbp will be easier to adapt to the plumbing architecture
18:57:27 <SumitNaiksatam> igordcard_: oh okay, let us know if you need any help with discussions or implementation
18:57:52 <igordcard_> SumitNaiksatam: I will :)
18:58:01 <SumitNaiksatam> igordcard_: thanks
18:58:07 <SumitNaiksatam> anything else we want to discuss today?
18:59:21 <SumitNaiksatam> alrighty, thanks everyone!
18:59:23 <SumitNaiksatam> bye
18:59:26 <ivar-lazzaro> bye!
18:59:36 <rkukura> bye
18:59:43 <SumitNaiksatam> #endmeeting