Which is better way for accessing berkeley db

I am using two package 1) Berkeley db Java Edition using BASE API 2)Berkeley DB Java E开发者_如何学Pythondition Using DPL (Direct Persistence Layer)

if the performance is the issue then which is better way

DPL is basically the base API with BDBs built-in custom binding/serialization on top.

This means that performance will depend on where the data you store is coming from. If it is coming from Java objects and you want to use the base API, it's unlikely that you'll find a quicker serialization process than BDBs. (Although not impossible and you can always construct one for specific cases.)

If the data you want to store is already in a serialized format, or just not in the form of POJOs, you might be better off with the base API.

Either way, the official recommendation is that unless you have a very good reason not to use it, you sdhould use DPL. And generally speaking the perfomance of databases depends on the number of times disk access is needed and that isn't affected much by the API you use.