Whether I use RocksDB or FS State backends, if my requirements are to have fault-tolerance and ability to recover with 'at-least once' semantics for my Flink job, is there still a valid case for using a backing local FS for storing states? i.e. If a Flink Node is invalidated, I would have thought that the only way it could recover (by re-starting the task on different node), is if the state is stored in a shared file system such as HDFS, S3 etc....
I am asking since I want to know if HDFS is a must have for my deployment. Thanks, Hayden
