I have two questions related to failover in Flume. 1) If the source is not accepting events because of a failure condition, is there a failover mechanism in Flume other configuring 2 agents to listen to the events, in which I will get duplicates.
2) I configured a sink processor for failover with 2 sinks, one logger and another HDFS. If the HDFS sink starts failing (say HDFS is no longer up), the events are not processed by logger sink. Is this excepted behavior. Is the failover only works with 2 avro sinks and not in the manner I have configured? I will appreciate it if somebody can point me in the right direction. -Mayur
