I am developing a REST like API in which an initial log in call will be made, a database row created, and a “log in session key”(GUID/uniqueidentifier) will be returned to the client. This key will then be used on all subsequent calls to the API as a security check, until log-out. On EVERY API call I plan to ask the database to look up the row using the key and, if the row’s time stamp has not yet expired, allow the API to serve what it needs to.
For a simple select statement that may happen hundreds of times per “log in session,” would pure SQL preform better than LinqToSQL in this scenario?
Pure SQL should perform better than LinqToSQL (if anything because it doesn’t need to construct and cache the query the first time), however, it depends on the number of users, the server capabilities, and how fast those “hundreds of times” happen if you should be concerned about it.