cancel
Showing results for 
Search instead for 
Did you mean: 

ABAP+JAVA in Cluster installation

Former Member
0 Kudos

Hello gurus,

I've read all tha SAP HA documents and have noticed that in several cluster installation they split up

the ABAP SCS and JAVA SCS from the CI when installaing a double-stack system (ABAP+JAVA).

As much as I know, the SCS instances doesn't require a lot of memory or cpu, they are not the problem when there is a bottleneck (correct me if I'm worng ...).

So, what is the reason for splitting up the SCS isntances from the cluster group of central instance ?

Please advice,

Dimitry Haritonov

Accepted Solutions (0)

Answers (1)

Answers (1)

Former Member
0 Kudos

Hello Dimitry,

Basically the SCS instances determine the CI. Without having running message and enqueue servers, there is not a running NetWeaver cluster at all. So these processes together with the database and the global file system are critical for the cluster and its high availability. These are the so called single points of failures and need to be protected by the failover cluster. For more details refer to sections 2.2.4, 2.3.2 and 4.1 of "SAP WAS in Switchover Environments" technology handbook available at service marketplace.

You are right that SCS processes do not consume a lot of memory and cpu. They do not require a lot of time to start so for performance reasons it makes sense to split them from the ABAP and Java processes, which are much slower. If the SCS instances are installed in a separate cluster group than the whole CI, the switchover process will take less time.

I hope I answered to your question.

Best Regards

Vyara

Former Member
0 Kudos

Dear vyara,

Thanks for the information. But as far as I know in double

stack system the ABA+JAVA has a separate resource from the SCS,

but they are in the same group.

So as I see it, there is a cluster only between DB and SAP system, I mean there is no clustering between ABAP, JAVA or SCS.

correct me if I'm wrong.

Best regards,

Dimitry Haritonov