17:09:59 <tbachman> #startmeeting tws
17:09:59 <odl_meetbot> Meeting started Mon Aug 25 17:09:59 2014 UTC.  The chair is tbachman. Information about MeetBot at http://ci.openstack.org/meetbot.html.
17:09:59 <odl_meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:09:59 <odl_meetbot> The meeting name has been set to 'tws'
17:10:15 <networkstatic> ty thomas lol
17:10:18 <tbachman> #topic Michael Rehak covering OpenFlow plugin’s new exteensions
17:10:21 <tbachman> np!
17:11:23 <tbachman> folks — missed the first parts (multi-tasking). If anyone followed what Michael covered, could you #info?
17:13:07 <tbachman> #info recommend looking at the recording, as Michal Rehak was showing the architecture using a diagram from his (shared) desktop
17:13:09 <networkstatic> #link https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Backlog:Extensibility Michal R. is reviewing openflowplugin project extensnsibility
17:13:26 <tbachman> networkstatic: thx!
17:13:32 <networkstatic> thank u sir
17:16:49 <tbachman> #chair networkstatic
17:16:49 <odl_meetbot> Current chairs: networkstatic tbachman
17:17:14 * tbachman handing networkstatic meetbot powers :)
17:18:03 <tbachman> #info questions about where the extensions will live, and who wil maintain them. networkstatic wanted to make sure there was ownership.
17:18:21 <tbachman> #info edwarnicke says that this won’t change for Helium, given that M5 (code freeze) is 1 week away.
17:22:58 <tbachman> #info networkstatic asks where there are examples of implementatino
17:24:02 <tbachman> #info Michal provided some examples on the wiki
17:25:50 <tbachman> #info Michal says that this is currently only tested in the OpenFlow plugin distributions
17:29:06 <networkstatic> #link https://git.opendaylight.org/gerrit/gitweb?p=ovsdb.git;a=commit;h=50c3b5e41ce355dd4603a79004f2980a2f7cb5f1 Example Java API extension usage
17:30:25 <networkstatic> #link https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Backlog:Extensibility#Helpful_hints_for_reviewing_https:.2F.2Fgit.opendaylight.org.2Fgerrit.2F.23.2Fc.2F9393.2F Example robot framework and RESTconf usage
17:36:49 <tbachman> #info Michal says that the result from the RPC only provides the result of sending. It doesn’t provide the result of the complete transaction. An error listener should be used to catch the full result, which are sent as notifications
17:38:40 <tbachman> #info The XID in the notification matches the XID of the original RPC
17:40:16 <networkstatic> #link https://gist.github.com/a7b792e57e694ea45774 debating what a Barrier is https://gist.github.com/a7b792e57e694ea45774
17:45:49 <tbachman> time check — was there another topic to cover?
17:47:43 <cdub> you'll get a response when you get it...all you know is the barrier completed when it completes
17:50:36 <networkstatic> abhijitkumbhare https://bugs.opendaylight.org/show_bug.cgi?id=1504
17:50:59 <networkstatic> nevermind abhijitkumbhare
17:51:01 * tbachman wonders if there’s a barrier to this thread ;)
17:51:10 <networkstatic> its tied to a bug Dave tucker has
17:51:17 <networkstatic> haha tbachman
17:51:41 <cdub> tbachman: that won't make it stop, just order what happened before and next
17:51:47 <tbachman> cdub: ROFL
17:51:53 <abhijitkumbhare> cdub :)
17:52:01 <cdub> where next starting to look like, there is no next, ETIMEDOUT
17:52:24 <networkstatic> udp vs tcp :)
17:52:36 <networkstatic> ish
17:52:40 <Madhu> sorry guys. tbachman cdub
17:52:45 <Madhu> it was top of my mind to get this solved
17:52:50 <Madhu> today we are ERROR free :)
17:52:58 <tbachman> Madhu: no worries - just keeping it light :)
17:53:06 <Madhu> because we don't know how to handle it :(
17:53:16 <cdub> Madhu: i don't mind, just joking w/ tbachman
17:53:49 <Madhu> I know :) .... but it gets very tricky to implement this at the application level
17:54:09 <Madhu> because every call we make... it is difficult to maintain states for indefinite time
17:54:28 <tbachman> #info Is that an application consideration tho?
17:54:36 <tbachman> (i.e. the app knows what it should do in that event)
17:55:00 <tbachman> (and thus determine a wait time)
17:55:42 <cdub> normally, you'd issue a barrier when the operation you want to do depends on results from earlier operation, and rest of state is updated via async callback as async operations complete
17:56:10 <cdub> iow, order matters
18:00:51 <edwarnicke> Apologies all... I have to drop
18:01:04 <tbachman> edwarnicke: no worries — busy guy! :)
18:13:37 <networkstatic> #link https://www.wireshark.org/download/automated/ FYI, Wireshark has started getting Nicira Extension dissector support in the nightly builds
18:16:21 <Madhu> #info Stats with extensions seems to have problem (we saw preconditions errors during the error)
18:19:04 <networkstatic> #link not sure if this is related or not, https://bugs.opendaylight.org/show_bug.cgi?id=1484
18:26:32 <tbachman> #endmeeting