I don't want to seem rude, but this is the sort of question the briefest search would have answered. It's always worth doing a bit of hunting for an answer before posting a question; you'll probably get your answer much quicker.
The N+1 problem exists where a coder takes a somewhat naive approach to accessing relational data. Imagine you have a data access object to manage a User and another one to manage their Roles and these map to two tables (user and role) with a one to many relationship between them. You could have a method called getUsers that queries for users based on some criteria. As part of that method it could also get the Roles for each user from the other DAO. This is where the n+1 problem occurs, because you would have one query to get the Users the n queries to gets roles for each user (e.g. by running select *from roles where user_id = ? ). In pseudo code it would look something like this:
People more used to working in relational databases will spot that this could be better done with one query, probably by using a join to the roles table in the first query.