16:06:28 <moizr> #startmeeting “clusteringhackers” 16:06:28 <odl_meetbot> Meeting started Tue Nov 18 16:06:28 2014 UTC. The chair is moizr. Information about MeetBot at http://ci.openstack.org/meetbot.html. 16:06:28 <odl_meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:06:28 <odl_meetbot> The meeting name has been set to '_clusteringhackers_' 16:06:36 <colindixon> #topic agenda bashing 16:06:45 <tbachman> :) 16:06:47 <moizr> #chair tbachman 16:06:47 <odl_meetbot> Current chairs: moizr tbachman 16:07:01 * tbachman needs to learn to type in the right IRC channel ;) 16:07:04 <colindixon> somebody other than me should #topic it, since I can’t 16:07:10 <tbachman> #topic agenda bashing 16:07:17 <tbachman> #info two node clustering update 16:07:28 <tbachman> #info testing (integration) status 16:07:33 <tbachman> #info other items 16:07:59 <tbachman> #info colindixon asks the status of migrating the existing core applications (e.g. openflowplugin) 16:08:10 <tbachman> #info moizr says there’s no explicit work going on in that area 16:08:29 <tbachman> #info moizr says probably not all the apps are designed to support clustering 16:08:47 <tbachman> #info jmedved says there is work going on in the openflowplugin to help make it clustering aware 16:09:39 <tbachman> #info moizr would like to see guidelines on how projects should change their apps to support clustering 16:10:04 <tbachman> #info colindixon says we have best practices: 1) store everything in the MD-SAL; 2) do app local caching where needed 16:10:57 <tbachman> #topic bugzilla review 16:11:15 <tbachman> #info moizr says we had about 29 last week, now down to 17, most related to performance enhancements 16:11:41 <tbachman> #info all the integration tests for 3-node cluster are now working 16:12:35 <tbachman> #info The link state topology issue was discovered with BGP, where serialization of binary data was a problem 16:14:09 <tbachman> #info The in-progress bugs were suspended while getting the SR1 bugs fixed 16:15:49 <tbachman> #info There only bug in the confirmed status that needs to be addressed is the one where the leader should abandon (Bug ID 2371) 16:17:12 <tbachman> #info pantelis asks about the serialized UID version in some of the RAFT messages to get rid of sonar warnings, and whether these will be backed out (breaks backwards compatibility) 16:18:07 <tbachman> #action pantelis will create a bug to keep track of the serialized UID version in the RAFT messages 16:18:21 <tbachman> #topic two node clustering update 16:18:28 <markmozolewski2> https://git.opendaylight.org/gerrit/#/c/12832/ 16:18:44 <tbachman> #link https://git.opendaylight.org/gerrit/#/c/12832/ gerrit review for current two-node work 16:19:16 <tbachman> #info markmozolewski2 says there are some additional changes that will be empty unit tests to cover the use cases that they plan to support 16:19:37 <tbachman> #info markmozolewski2 says they also rely on a network partition detection algorithm that’s separate from this code. 16:19:56 <tbachman> #info markmozolewski2 requests folks to have a look and provide comments on the gerrit 16:20:03 <tbachman> #topic testing (integration) status 16:20:34 <tbachman> #info moizr says we still haven’t had a stable build on the integration branch 16:20:59 <tbachman> #info moizr checked clustering, and says that all the tests on single-node openflow are now passing 16:22:56 <tbachman> #link https://git.opendaylight.org/gerrit/#/c/12831 gerrit for refactoring clustering integration tests 16:23:15 <tbachman> #info moizr would like to see one other person try this out 16:23:39 <tbachman> #info etka says they are 80% of the way of getting the 3-node setup 16:23:55 <tbachman> #info catohornet says she can get to the clustering test by the end of the week 16:24:14 <tbachman> #info moizr says all the VMs need the same username and password for the 3-cluster test 16:26:42 <tbachman> #endmeeting