It will help better if you provide the test scenarios/cases that you wish to cover and any background information on them. Also provide information on how you to intend to test them (or have tested for them).
From what you are indicating, it seems that one of your instances failed and only THAT instanced failed over to the standy instance. With your configuration, this seems to be ok (but it seems you were expecting ALL instances to failover).
Remember that "HA" is a very generic term - and can mean different things to different people (e.g. for Apps Arhitect it means his "browser" session should be tolerant to anything happening on the infrastructue). In the realm of DB2, HA can be implemented at the OS level (using clustering s/w or using DB2 HADR technology similar to Oracle's DataGuard). It seems you are using HADR.
The more info you provide the better help we can provide.
first of all thnx for ur response then
i'll rebuild my Q? i need to know only if there exist case that it could happen that only one instance failed and failed over and the others remains up and running (e.g. in my sketch i've nodeA with db2inst1 , ..3 ,..5 active and nodeB with db2inst2,..4 active is there exist such a case that db2inst2 in nodeB fails and failed over on nodeA and db2inst4 in nodeB remains up and running in nodeB)