Of course that is the case, Oracle RAC consists of several nodes (in our case usualy 8 nodes/RAC)

What I am after is a way to automaticaly in TD identify Oracle RAC, bounding them together with the TNS listeners and the ASM with for example the "service name" of the RAC.

For normal stand alone databse instaces, applications use hostname, instance name, port to connect to the databse.

For Oracle RAC applications connect to the "service name" of the RAC that consists of all nodes.

Do I need a modification of how Discovery collects the db information by the TNS method? If so, any idea of what has to be modified? Or any other suggestions or thoughts at all?

This question and ellaboration has been idle long enough...

I´m looking in to doing this with UAD, similar to the "Oracle DB (specialized)" entry in the Application Library. Probably it would be necessary to some work with the "Oracle Listener (specialized)" entry in the Application Library as well if not creating a new entry for it. Have not looked in to that specific part yet.

The processes I´m working with are:

racgimon (with keyword "asm") (specifies an instance)

asm_ckpt

asm_dbw0

asm_diag

asm_gmon

asm_lck0

asm_lgwr

asm_lmd0

asm_lmon

asm_lms0

asm_mman

asm_pmon (specifies an instace)

asm_psp0

asm_rbal

asm_smon

Still have some rather big work with "Regular Expressions" to get it all sorted.

For the Listener I´m not sure how to work with this yet...

But in a perfect world it would be great to identify all hosts, instances and nodes of the RAC, the listener ServiceName and the clustername of the RAC.

I´ll keep you posted, input or comments always welcome