The Most Scalable EJB Idiom

However, I just stumbled upon an EJB Benchmark put together by UrbanCode (the makers of AntHill). It's a bit dated, that is almost a year ago, however I suspect the numbers remain applicable even today.

They essentially go about testing different EJB idioms:

1. Fine Grained Entity Bean (FGE)
   often used by beginners to Enterprise Java Beans
2. Coarse Grained Entity Bean (CGE)
   the best performing Entity Bean idiom typically used in practice
3. Coarse Grained Container Managed Entity Bean (CGCME)
   touted as faster then Bean Managed Persistent Entity Beans
4. Optimized Entity Bean (OE)
   the best performing contender that still uses Entity Beans
5. Session over Entity Bean (SE)
   typically advanced as a "best practice" in trade media
6. Coarse Grained Session Beans (CGS)
   the proposed best performer

to eventually conclude the obvious. The most scalable idiom is the one that uses zero (i.e. none, zippo, zilch) Entity Beans.

Not really surprising, however their benchmark (also downloadable) gives me hard evidence to back my previous assertions.

So if any of these "Enterprise Architect" types try to talk you into an insane architecture, you now have serious ammunition to prove them wrong. Tough to argue against real code and solid numbers.