Tasks

IDEA: Detect N + 1 check session factory stats

		if( "statements closed" - 
			( "entities loaded" + "entities updated" + "entities deleted: ) > "entities loaded" )
		{ N + 1 ! }             !!! Confirmed !!! :)
		This formula seems to be logically equivalent to "entities fetched" > 0
		as sessionFactoryStatistics.logSummary(); mentions ..."- entities fetched (minimize this)" it seems
		that this property counts the number of additional fetches performed when an object is loaded.
		
		Question ? :
		Does "entities fetched" > 0 indicate the presence of an n + 1 problem since it is the number of secondary loads.
		If so is this true when n = 1 ?
		
		If this is true then
		If the stats data was cleared ( or sessionFactory created ) and a query run ( this to associate a query with its results ),
		then a non zero value for "entities fetched" at that point would
		indicate an n + 1.  If this procedure were added to unit tests it could be used as an automated check for n + 1
		problems in a systems data access.     
		

TASKS:

Fully automate tests by capturing relevant sql and parameterising n

Figure out relationship between join types and annotations

Test N + 1 detection by adding it to test cases and compare to results
Determine which query caused the n + 1

Complete associations matrices for all association types

Check em.find() and equivalent queries for n + 1 behaviour, is n = 1 in this case ?

Mark annotation default options in Person and Address

Check byte code interception effect on nullability caused n + 1 scenenarios  

Check cascading fetch scenarios A -> B -> C ( would expect to see statement complexity = A * B * C e.g. does an n+1 at ->AB and ->BC = n * n fetches ? )

Check parallel n + 1 scenarios i.e. A -> B, A -> C e.g can a n+1 at ->AB and ->AC be resolved with join fetching

Convert HQL queries to named queries 

Check join fetch solutions in criteria form

L1 Cache
	
	Examine L1 cache use on fetch of a already fetched object	
	
	Examine L1 cache use on lazy association fetch of a already fetched object 

	Examine L1 cache use on query fetch by matching condition of a already fetched object

Examine Setting fetch depth