They are not the same, but I don't blame you for thinking that they are. The reason why it doesn't seem clear that they are the same is that you've only seen one example of each.
So let's step back, define them separately, and then look at some interesting examples.
Propositional logic is a branch of mathematics that studies propositions, their truth or falsity, and how they combine.
What you probably think of as "propositional logic" is actually just one kind of propositional logic, namely, classical logic. However, this is not the only kind of classical logic and not the only one that is interesting in computer science.
Some theories are built on top of classical logic. Presburger arithmetic, for example, is the theory of natural numbers with addition. Tarski arithmetic is the theory of real closed fields. First-order logic is the theory of quantified variables over non-logical objects.
You can think of these logic systems as types of propositional logic, with more axioms to deal with the extra stuff. Many constraint-type problems are expressible in this framework; you can look up satisfiability modulo theories for more information on how this is used.
But there are also propositional logics that are not "classical". One example of a non-classical logic is intuitionistic logic. This is a logic that models constructive proofs, which is interesting because constructive proofs are the same thing as well-behaved computer programs. ("Well-behaved" in this sense means there is no unbounded recursion.) They are also the basis for modern type systems, such as that of Haskell.
Constructive proofs (e.g. pick any proof from Euclid's Elements) are implementable as computer programs. Given some geometric objects as input, you could implement the proof (at least in principle) as a computer program that performs the construction.
You can't do that with non-constructive proofs. Let's look at an example to see what I mean. I'll prove that there are two irrational numbers, $a$ and $b$, such that $a^b$ is rational.
Consider $z = \sqrt{2}^\sqrt{2}$. This number is either rational or irrational.
Case 1: If $z$ is rational, then set $a = b = \sqrt{2}$, and the theorem is proven.
Case 2: If $z$ is irrational, set $a = z$ and $b = \sqrt{2}$. Then $a^b = 2$ and the theorem is proven.
QED
The reason why this proof isn't constructive is that it relies on the law of the excluded middle; $z$ must be either rational or irrational, but we don't (and, in a sense, can't) actually test which of them is true.
Intuitionistic logic systems tend to give you proof systems that are decidable. This is an extremely important practical concern. If you consider, for example, the Java bytecode verifier, or the type system for a programming language (both of which prove that some program is type-correct), they must terminate with an answer, so they must be based on a decidable proof system. By allowing propositions that are neither provable nor refutable, you avoid Gödel-like incompleteness.
That's just one example, of course. Other non-classical propositional logics include modal logic (which can be used to model situations where you have multiple agents who know different things; think about network protocols), and temporal logics (where the truth or falsity of a proposition may change over time).
The thing that connects all of them is the notion of a "proposition". Propositional logic can be thought of as the study of a family of logic systems, all of which deal with the notion of a mathematical statement that has a "truth"-type judgment.
Boolean algebra, on the other hand, is a purely algebraic system, characterised by a set of axioms. Saying "Boolean algebra" is like saying "group" or "field" in abstract algebra. Classical propositional logic satisfies the axioms, but so do other mathematical structures.
The subsets of a set $S$, for example, form a Boolean algebra; with the "not" operator corresponding to "set complement", "and" corresponding to "set intersection", and "or" corresponding to "set union". You can think of "true" as being $S$ and "false" as the empty set, but these are not the only "truth values" which the system models.
Lattices form Boolean algebras, and some lattices are important in computer science (e.g. Scott domains). Moreover, some important non-classical logic-like systems naturally form Boolean algebras, such as fuzzy logic.
So what I hope you can see is that while the two notions are related (some Boolean algebras model some propositional logics), they are not exactly the same thing. How they differ is part of what makes them interesting.