Actually I think I figured out the issue with l2switch basic csit: For some reason after we changed to new OF plugin (old plugin is fine) the l2switch app does not work very well when mininet disconnect + reconnect with no time in-between.
To probe this I added some sleep (2 secs) between disconnect and connect mininet between tests, and after that the tests pass with no issues: https://git.opendaylight.org/gerrit/#/c/44799/ https://jenkins.opendaylight.org/sandbox/job/l2switch-csit-1node-switch-only-boron/5/ So I opened this bug to track the new behavior: https://bugs.opendaylight.org/show_bug.cgi?id=6575 I am not sure how critical is this as most users trying l2switch manually will not notice anything. BR/Luis > On Aug 29, 2016, at 10:30 AM, Luis Gomez <[email protected]> wrote: > > Hi l2switch devs, > > Here is my analysis on CSIT failures: > > - Address Tracking and Loop Removal test issues are not reproducible outside > ODL infra, therefore these are not blocker and maybe not even bugs. Still > some infra investigation is required here as same tests are passing for > Beryllium. > > - Issue with host scalability is due to recent regression: l2switch does not > push flood flows on single switch: > https://bugs.opendaylight.org/show_bug.cgi?id=6573 > > BR/Luis > > _______________________________________________ openflowplugin-dev mailing list [email protected] https://lists.opendaylight.org/mailman/listinfo/openflowplugin-dev
