Previously the same file name was used for each atomic action, causing overwriting.
Changing the logic to write to a different file each time.
+ Dump JSON in the relevant results directory
+ Update Elasticsearch indexing status in summary
+ Add failover to Shaker results also
Sample output: https://gist.github.com/smalleni/c25ef05c9b815a8e3299c2fc904908ef
Change-Id: Ib74cfd2ecf5c63857548c0ef219e7965b6323d56
e06f895455