====================================== #opendaylight-meeting: kernel projects ====================================== Meeting started by rgoulding at 16:01:31 UTC. The full logs are available at http://meetings.opendaylight.org/opendaylight-meeting/2018/kernel_projects/opendaylight-meeting-kernel_projects.2018-04-17-16.01.log.html . Meeting summary --------------- * agenda bashing (rgoulding, 16:01:37) * odlparent 3.1.0 and yangtools 2.0.3 upgrade status (rgoulding, 16:01:58) * increase in frequency in pax-exam timeouts (rgoulding, 16:04:12) * increase in frequency in pax-exam timeouts (rgoulding, 16:04:28) * LINK: https://git.opendaylight.org/gerrit/#/c/71041/ (rgoulding, 16:04:50) * vorburger: in the pax-exam tests you can specify the base karaf that you want to deploy into, and asks what the target is (rgoulding, 16:05:44) * these tests are set to use opendaylight-karaf-empty, which should be already converted to equinox (rgoulding, 16:06:13) * are we overriding it in the controller poms? (rgoulding, 16:07:49) * Exception— “ ServiceLookupException: gave up waiting for service org.ops4j.pax.exam.ProbeInvoker" (rgoulding, 16:12:21) * this used to happen but is happening much more frequently now (rgoulding, 16:12:34) * rovarga there is probably a race and the race window opened (rgoulding, 16:12:48) * this doesn’t reproduce locally (rgoulding, 16:12:56) * either this is reproducible locally or we need to retain the appropriate logs from jenkins to properly investigate (rgoulding, 16:13:58) * the appropriate information is not put into collected surefire logs (rgoulding, 16:14:57) * binding ITs are not using surefire, they are using failsafe (rgoulding, 16:15:47) * the toaster tests are using failsafe not surefire (rgoulding, 16:16:02) * fire off an email to integration to save off the correct information if test timeout after 2 hours (rgoulding, 16:19:27) * rovarga was able to reproduce locally (rgoulding, 16:20:58) * rovarga has logs and can ship over karaf.log (rgoulding, 16:22:12) * LINK: https://lists.opendaylight.org/pipermail/controller-dev/2018-April/014350.html (rgoulding, 16:22:15) * ^^ mail for this issue (rgoulding, 16:22:21) * sal-binding-it is pulling in jersey-core, and the same issues are happening that happened with AAA (rgoulding, 16:22:59) * LINK: https://git.opendaylight.org/gerrit/#/c/70979/ (rgoulding, 16:23:43) * ^^ fix in AAA (rgoulding, 16:23:48) * pax-exam is getting messed up if the feature service has trouble getting deployed (rgoulding, 16:26:50) * ACTION: tpantelis to try to remove jersey-core dependency (optional) to see if it is actually needed for the tests (rgoulding, 16:40:50) * this might have been used in formatting controller-config.xml, and may not be needed anymore (rgoulding, 16:42:19) * web centralization (rgoulding, 16:45:43) * we are changing api version so we need to recompile every consumer to make sure this would work. we would need to create an adaptor of javax.ws.rs-api in order to do this in a “single switch” (rgoulding, 16:49:18) * odlparent 3.1.0 and yangtools 2.0.3 upgrade (rgoulding, 16:52:00) * stable/oxygen was upgraded to odlparent 3.1.0 to get karaf upgrade (rgoulding, 16:53:24) * upgrade patches have been merged (rgoulding, 16:54:06) Meeting ended at 16:54:35 UTC. People present (lines said) --------------------------- * rgoulding (34) * odl_meetbot (3) * skitt (1) Generated by `MeetBot`_ 0.1.4