First round had too many questions regarding distributed systems so it's not possible to mention everything here. But mainly details around Apache Hadoop HDFS part and how it's built internally. Some questions around CAP and how it's achieved in HDFS + how it works internally and past experiences. Second interviewer came unprepared and asked me to implement two wrapper methods around HashMap (one to fetch and one to delete) was very useless question which can be solved by a CS student.