Run in this order:
sbt "run leader"
, wait forWaiting for cluster to form up...
log messagesbt "run worker"
, it will choke on one of the first messages- kill leader !important!
- run
jps | grep ClusterApp | cut -d ' ' -f 1 | xargs -I {} jstack {} | grep bomb -B 9 -A 23
to see thread hanging on CountDownLatch#await() - add
@transient
tolazy val bomb: String =
- repeat steps from 1 to 2 and check that worker node is now able to process all messages