Technical Guide: Setting Up SAP HANA System Replication (HSR) on AWS

By Kourosh Ghouchkhani – SAP Basis Cloud Architect

Overview

This guide outlines the steps required to set up SAP HANA System Replication (HSR) in an AWS cloud environment for high availability (HA) and disaster recovery (DR). HSR ensures data availability by replicating data from the primary HANA system to a secondary system in real time.

Target Audience

• SAP Basis Administrators

• Cloud Architects

• SAP Infrastructure Engineers

• Enterprise Architects managing HANA on AWS

Prerequisites

AWS Infrastructure

• Two or more EC2 instances (for primary and secondary HANA nodes)

• Shared VPC or VPC peering between regions if cross-region DR is needed

• EBS volumes provisioned for /hana/data, /hana/log, /hana/shared

• SAP-certified EC2 instance types (e.g., r6id.16xlarge)

SAP HANA

• HANA 2.0 SP05 or higher

• HANA installed and configured on both primary and secondary nodes

• Matching OS versions and kernel patch levels

• Hostname and SID must be the same on both systems

• Time synchronization via NTP

Security & Access

• AWS Security Groups with ports open for replication (default TCP 3xx13)

• SSH key exchange between nodes

• Appropriate IAM role and permissions for EC2 and storage

Architecture

pgsql

CopyEdit

+---------------------+ +---------------------+ | Primary EC2 (AZ-A) | <====HSR====> | Secondary EC2 (AZ-B)| | SID: PRD | Replication  | SID: PRD | | IP: 10.0.1.10      | | IP: 10.0.2.10       | +---------------------+ +---------------------+ Active DB Standby DB /hana/data, /hana/log                 /hana/data, /hana/log

Optionally integrate with AWS Route 53, Elastic IP, and Pacemaker/Cluster Manager for full HA automation.

Step-by-Step Setup

1. Prepare Operating System and File Systems

• Ensure /etc/hosts is updated with both HANA hosts.

• Set up LVM or EBS volumes for required mount points.

• Match UID/GID and permissions across nodes.

bash

CopyEdit

mkdir -p /hana/data/PRD /hana/log/PRD /hana/shared /usr/sap/PRD

2. Install SAP HANA on Both Nodes

Use the same installation media and SID, and configure as standalone on both.

bash

CopyEdit

./hdblcm --action=install --sid=PRD --number=00

3. Enable System Replication on Primary

Log in as sidadm on the primary node.

bash

CopyEdit

hdbnsutil -sr_enable --name=primary

Check replication status:

bash

CopyEdit

hdbnsutil -sr_state

4. Register Secondary Node

On the secondary node, register the system:

bash

CopyEdit

hdbnsutil -sr_register --name=secondary --remoteHost=10.0.1.10 \ --remoteInstance=00 --replicationMode=sync --operationMode=logreplay

5. Verify HSR Status

On both nodes:

bash

CopyEdit

hdbnsutil -sr_state

You should see:

• ACTIVE on the primary

• SYNC or SYNCMIRROR on the secondary

6. Configure Automatic Takeover (Optional)

Enable auto failover only when using clustering solutions like Pacemaker or AWS Failover Scripts:

bash

CopyEdit

sapcontrol -nr 00 -function EnableAutoFailover 1

Post-Configuration

• Set up monitoring with SAP HANA Cockpit or AWS CloudWatch

• Implement backup using AWS Backup or HANA Studio

• Periodically test failover using:

bash

CopyEdit

hdbnsutil -sr_takeover

• Harden with encryption, IAM policies, and firewall rules

Optional Enhancements

• Route 53 DNS Failover – Point SAP application layer to a DNS CNAME

• SAP HANA Scale-Out with HSR – For large deployments

• Pacemaker Cluster – For automated HA with fencing

Final Notes

• Always test in a non-production environment.

• Ensure SAP Note compatibility with AWS services.

• Validate your replication setup regularly.

Need Help?

I offer consulting services for SAP HANA high availability on AWS, including:

• Architecture design

• RISE with SAP HSR implementation

• Cloud-native automation

• Performance tuning

Let’s connect to design a resilient, future-ready SAP landscape on AWS.