While playing around locally while starting to learn OpenStack with OpenDaylight using my https://github.com/vorburger/openstack-devstack-opendaylight-vagrant/ I've managed to get ODL into a state where it would never stop printing this WARN, every 3s, and thus clogging the log with it...

Not sure exactly how I got it into that state (I think I stopped my VM running devstack), but whatever the reason, it seems to me that this "statistics collection" could be made a bit more robust, and if it fails (maybe not immediately but after a number of times?) it could "give up", maybe print an ERR instead of WARN for operator FYI, and stop doing this?

2017-03-09 11:34:30,846 | WARN | pool-33-thread-1 | StatisticsManagerImpl | 314 - org.opendaylight.openflowplugin.impl - 0.4.0.SNAPSHOT | Unexpected error occurred during statistics collection for node openflow:119556424750016, rescheduling statistics collections
java.lang.Throwable: Device connection doesn't exist anymore. Primary connection status : RIP
at org.opendaylight.openflowplugin.impl.statistics.StatisticsContextImpl.deviceConnectionCheck(StatisticsContextImpl.java:325)
at org.opendaylight.openflowplugin.impl.statistics.StatisticsContextImpl.gatherDynamicData(StatisticsContextImpl.java:156)
at org.opendaylight.openflowplugin.impl.statistics.StatisticsContextImpl.gatherDynamicData(StatisticsContextImpl.java:147)
at org.opendaylight.openflowplugin.impl.statistics.StatisticsManagerImpl.pollStatistics(StatisticsManagerImpl.java:157)
at org.opendaylight.openflowplugin.impl.statistics.StatisticsManagerImpl.lambda$scheduleNextPolling$1(StatisticsManagerImpl.java:213)
at io.netty.util.HashedWheelTimer$HashedWheelTimeout.expire(HashedWheelTimer.java:625)[146:io.netty.common:4.1.8.Final]
at io.netty.util.HashedWheelTimer$HashedWheelBucket.expireTimeouts(HashedWheelTimer.java:700)[146:io.netty.common:4.1.8.Final]
at io.netty.util.HashedWheelTimer$Worker.run(HashedWheelTimer.java:428)[146:io.netty.common:4.1.8.Final]
at java.lang.Thread.run(Thread.java:745)[:1.8.0_121]