generic vs. specific userID in PI Interfaces

[Former Member]

Hi All

We have many interfaces that are running between R3 and PI. Due to recent performance problems in R3 system, Basis team wants to differentiate between which application is causing and suggested to create Interface specific user id vs ALEMASTER which is used now.

Can anyone suggest if this is best practice to use PI Interface/application specific user id in PI channels or use a one channel with ALEMASTER. I know we have to create that many RFC destination and PI channels...which might bring down performance of the system connectivity.

Also for RFC channels we use logon groups with specific application servers on it.

Any inputs on this is appreciated.

Thanks,

Giridhar.

[MichalKrawczyk · Accepted solution]

Hi,

>>>Basis team wants to differentiate between which application is causing and suggested to create Interface specific user id vs ALEMASTER which is used now.

this is not an issue of PI user it's because:

- the receiving application don't have any source system identifiers

- the message which posts the transaction does not have any additional keys (like in IDOC function code etc.)

so don't change the user, change the error determination procedures in order to do it correctly

Regards,

Michal Krawczyk

[Former Member · Accepted solution]

Thanks for your answer.

Also can you please tell me the positives and negatives of using a single userID (ALEMASTER) instead of workstream-specific (or interface-specific) IDs for communication between PI and R3.

Thanks

Giridhar.