Many database professionals or even other people who have any concern with databases pronounce SQL as sequel, although the correct pronunciation of the language is S-Q-L ("ess kyoo ell"). According to my guess there are both historical reasons and linguistic resons behind it.
We need to look back into the history to know reasone. In the 1970s IBM developed a language called SEQUEL, which was an acronym for Structured English Query Language. This language was designed inorder to manipulate data stored in a database system known as System R. Later on, the acronym SEQUEL was shortened to SQL because of a trademark dispute.
In 1986 ANSI adopted SQL as a standard, and in 1987 ISO did so. ANSI declared that the official pronunciation of the language is "ess kyoo ell," but it seems that this fact is not common knowledge.
As far as linguistic reason is concern, the sequel can be pronunced more fluently, mainly for English speakers. I have to say that I often use it myself for this reason.
You can sometimes guess which pronunciation people use by inspecting their writings. Someone writing "an SQL Server" probably uses the correct pronunciation, while someone writing "a SQL Server" probably uses the incorrect one.
So that was the stroy behind SQL and SEQUEL.