Managing Serviceguard Extension for SAP, December 2007

Step-by-Step Cluster Conversion
SAP Preparation
Chapter 3106
Creation of Replication Instance
This section describes how to add Enqueue Replication Services (ERS) to
a system that has ASCS and/or SCS instances. The ASCS (ABAP) or SCS
(JAVA) instance will be accompanied with an ERS instance that
permanently keeps a mirror of the [A]SCS internal state and memory. As
a result, no information gets lost due to failing SAP System Central
Services, if the SGeSAP package that includes [A]SCS fails over to the
node that runs ERS during runtime.
NOTE If both SCS and ASCS are intended to take advantage of Replicated
Enqueue, then the following steps need to be taken twice.
RE070 Replicated Enqueue Conversion:
Create a new mountpoint:
su - <sid>adm
mkdir /usr/sap/<SID>/ERS<INSTNR>
RE080 Replicated Enqueue Conversion:
A volume group needs to be created for the ERS instance. The physical
device(s) should be created as LUN(s) on shared storage. Storage
connectivity is required from all nodes of the cluster that should be able
to run the Replicated Enqueue. For the volume group, one logical volume
should get configured.
Mount the new logical volume to the mountpoint created in RE070.
RE090 Replicated Enqueue Conversion:
Create instance subdirectories in the mounted logical volume. They will
be switched between the cluster nodes later.
su - <sid>adm
cd /usr/sap/<SID>/ERS<INSTNR>
mkdir data log exe work profile
Starting with SAP kernel 7.0, the following subdirectory structure also
needs to be created:
mkdir -p exe/servicehttp/sapmc