Updated Guidance for SAP Pacemaker configuration – ABAP SAP Central Services (ASCS)
SAP customers running critical workloads on Linux on AWS often rely on Pacemaker, an open source resource manager, to orchestrate maximum availability. Pacemaker detects and reacts to failure scenarios that impact SAP uptime by disrupting the underlying infrastructure, the operating system or the SAP application itself.
AWS worked with SUSE and Red Hat on the original resource agents and documentation required to support the deployment of SAP on AWS, and we continue to collaborate to improve reliability, and minimise outages. This blog intends to highlight some of the recent documentation changes for the setup of the ABAP SAP Central Services (ASCS), a single point of failure for SAP NetWeaver applications which controls user load distribution and locking.
The updates demonstrate how to take advantage of new pacemaker parameters and SAP integration options, expand on core concepts such as the Overlay IP and Fencing Agents, and provide a more comprehensive list of setup checks and failure test cases. To improve the flow of the documents, we have separated the guidance into two documents, aligned to whether you are performing the setup on SUSE or Red Hat.
We encourage administrators to regularly update, review and test their setups. Our experience shows that operational procedures and monitoring mechanisms which are well understood are as important as the software and configuration of Pacemaker in the protection of your systems.
The following changes are included in the documentation.
- Simple Mount Configuration Option (SUSE Only) – SUSE has certified a new ASCS and ERS cluster solution using Simple-Mount file system structure. It simplifies the NFS shared file system structure for ASCS and ERS, and relies on SYSTEMD configuration to manage filesystem dependencies. As a result, the solution is more robust than the classic configuration. This is now the default and recommended configuration for both NetWeaver ENSA1 and S/4HANA ENSA2. You can read more about simple-mount in the SUSE Blog – Simple Mount Structure for SAP Application Platform
- Fencing Priority – This option increases the likelihood that the node which survives a fence race will be the node which is hosting your most critical resources – in this case the ASCS Message Server and Enqueue Server. Fencing priority is only supported for ENSA2 configurations.
- Monitoring Options – With the recent Amazon CloudWatch Application Insights support of NetWeaver, it is possible to monitor your ASCS cluster. We have also provided a simple script which can integrate with SNS or AWS Eventbridge to notify you if a pacemaker action was performed. In cases where failover is undetected (due to time of day or lack of activity) this can help with the diagnosis of underlying issues.
To get started with configuring Pacemaker for the ASCS or to review your existing setup, check out the documentation at the following links:
- SAP Netweaver on AWS: SUSE Netweaver (ASCS) High Availability
- SAP Netweaver on AWS: Red Hat Netweaver (ASCS) High Availability
Join the SAP on AWS Discussion
In addition to your customer account team and AWS Support channels, AWS provides public Question & Answer forums on our re:Post Site. Our SAP Specialists regularly monitor the SAP on AWS topic and will try to assist with any queries related to this topic.
To learn why thousands of customers choose AWS for SAP, check out the AWS for SAP page.