According to the Hadoop YARN API, here is the breakdown of the Container ID string:
"The format is container_e*epoch*_*clusterTimestamp*_*appId*_*attemptId*_*containerId*.
When epoch is larger than 0 (e.g. container_e17_1410901177871_0001_01_000005). *epoch* is increased when RM restarts or fails over.
When epoch is 0, epoch is omitted (e.g. container_1410901177871_0001_01_000005)."
To add more context, this was added by YARN-2562 in Hadoop 2.6.0 as an improvement for readability of the containerId string. So, its an expected behavior.