- 1st Round - Write singleton class structure with multithreading scenarios in mind - Given binary tree with parents record in DB, write query to fetch root node, intermediate nodes and leaf nodes - Given 2D Matrix Array with 1s and 0s, fan out 1 on left, right, bottom, top if present in matrix and do it in O(n x m) time complexity without using Space of O(n x m) - Basics of Kafka , Partition, offsets, consumer groups and how Kafka is used as message broker - 2nd Round - Getting dump of 100 log files daily with 1 GB data each named as log1, log2...etc....given an instance of time of a day find which log file will have logs related to that timestamp in efficient way - Directory of phone numbers is given of billion records with duplicate entries how do you store it efficiently (Hint: Set and Trie) - 3rd Round - Existing Project Architecture Explanation - Modified ingestion to use real time data instead of once a time loading/update of data - Modify architecture - Basics of SQL, NoSQL DBs, Internals of Indexing, how data inside index tables is stored by DB Engines - Indexing really improves performance ? Composite Indexing is overhead or optimal ?
