SourceForge.net Logo

FAQ

This page tries to compile the most asked questions and to answer as directly as possible.

If your question is not listed please fill a bug report / feature request here.

what store mechanism should I choose ?

As any good answer, this one starts with "it depends" :)

Each storage options targets its own applications with specific (and different) requirements and characteristics. Thus, using either OJB or XML has impact on how BERSERK fulfills application's requisits.

OJB is a much more complete storage mechanism. It has great extensibility and time complexity is way smaller than XML's. This means that for many filters, services and filter chains OJB has a smaller configuration reading overhead that XML.

On the other hand XML is easier to use and has better portability, since it needs no DBMS to store BERSERK's components configurations. It's time-to-run is way smaller, because the filters, services and filter chains configurations are stored on regular files on the OS native file system.

When running BERSERK on a managed environment (as Tomcat, JBoss, etc...), it may be easier to use OJB if you want to make runtime changes to service, filters and filter chain configurations, since manipulating files in a managed evironment can introduce aditional complexity (for example with file paths).

The conclusion is:

Use XML for quick developing (and portability) and OJB for wider possibilities and better scalability.

my Filter Chain failed. How can I know which filter(s) actually failed ?

 
BERSERK V1.0 RC3 added the possibility to inspect the FilterChainFailedException in order to conclude on what filter(s) failed. To do this, you may access a field in the exception.
 
The field is an HashMap structure containing as many entries as Invocation Timings (by the release of RC3, PRE and POST). Each entry contains a List with all the exceptions thrown by the offending filters.
 
It is now possible to customize the filters' exception messages and get them through to the service caller.