Interview 1 - He had block diagram of a search engine. Search text is converted into tokens and those are store in the database (keywords, time, city). Another utility (data collector) retrieves the data DAILY to fetch top 10 search keywords for a particular city. search text -> tokenizer -> database database -> data collector -> utility to save top 10 search keywords Let's say the system was running fine for 3-4 months but suddenly it started giving timeout to data collector. What can be the reasons? Data volume? How to solve this? Distributed data store? Based on what (assuming we don't need data beyond 7 days)? Write some queries and then tell him on which fields do we need indexing on. Data structure for the utility to store top 10 search keywords for a city.