Thank you, indeed it was an internet issue, changing the dns resolver in "/etc/resolv.conf" worked. default was 127.0.0.11, but this kept spitting errors, changing it to 8.8.8.8 worked. not sure why though. ... View more

it is a single cluster environment, how can there be network issues? I did not change any configs, that is the weird thing. ... View more

Thank you, indeed it worked with javac -cp `hadoop classpath` -d wordcount_classes WordCount.java ... View more

I had this error too. Only solution I found was to shut down and reboot. Apparently sometimes when you shut down the spark context in zeppelin it still somehow persists and hangs. Workaround for me was to stop all services in ambari, reboot the system and start again. Then the session is flushed. (Found it here: https://stackoverflow.com/questions/35515120/why-does-sparkcontext-randomly-close-and-how-do-you-restart-it-from-zeppelin ) ... View more