16:01:50 <rovarga_> #startmeeting Kernel Projects Call
16:01:50 <odl_meetbot> Meeting started Tue Jun 11 16:01:50 2019 UTC.  The chair is rovarga_. Information about MeetBot at http://ci.openstack.org/meetbot.html.
16:01:50 <odl_meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:50 <odl_meetbot> The meeting name has been set to 'kernel_projects_call'
16:01:56 <jamoluhrsen> https://jira.opendaylight.org/browse/NETCONF-626
16:07:43 <rovarga_> #agenda
16:07:57 <rovarga_> #info NETCONF-626
16:08:02 <rovarga_> #info CONTROLLER-1900
16:08:51 <rovarga_> #topic NETCONF-626
16:08:58 <rovarga_> #link  https://jira.opendaylight.org/browse/NETCONF-626
16:11:29 <rovarga_> #info we fail to share SchemaContext when devices are being connected concurrently
16:14:54 <rovarga_> #info the caching is being done in SharedSchemaContextFactory in yangtools
16:19:33 <rovarga_> #info the problem is that while we check the cache before trying to assemble a SchemaContext, if we have concurrent access we do not recheck if a result was computed
16:19:49 <rovarga_> #info this results in classic TOCTOU race
16:20:21 <rovarga_> #info which is not harmfull (the system works) but is highly inefficient, as the concurrently-computed results are used independently aree not deduplicated
16:20:40 <rovarga_> #info once a result is loaded though, all subsequent accesses will use the last computed result
16:21:28 <rovarga_> #topic CONTROLLER-1900
16:21:34 <rovarga_> #info Fluorine SR3 blocker
16:21:42 <rovarga_> #info it seems we have the root cause
16:22:01 <rovarga_> #info it's not that bad, as it does not affect general CDS users
16:22:48 <rovarga_> #info but it is bad in that it affects EOS
16:24:13 <LuisGomez> #link https://jenkins.opendaylight.org/releng/job/controller-csit-3node-clustering-ask-all-fluorine/237
16:26:43 <rovarga_> #info at the end of the day it is an issue where latest state may not be visible on followers for up to heartbeat interval
16:27:09 <rovarga_> #link https://jira.opendaylight.org/browse/CONTROLLER-1900
16:27:38 <rovarga_> #info this is hurtful when the selected EOS owner is on a follower, as its mechanics are tied to DTCL on that follower
16:27:55 <rovarga_> #info hence there can be a delay until the selected candidate becomes operational
16:28:14 <rovarga_> #info we have a patch, being validated at the link Luis linked
16:28:23 <rovarga_> #info which should alleviate the problem
16:48:50 <rovarga_> #endmeeting