Migrating SAP HANA to AWS: Tips and Tricks

SAP HANA is an in-memory relational database that provides real-time data processing and high-performance analytics. Amazon Web Service (AWS) services enable you to deploy SAP HANA in a highly available, fault-tolerant, and cost-effective way. When deploying SAP HANA on AWS, you get the performance benefits of SAP HANA along with the flexibility and scalability of the cloud.

Why Use SAP HANA in the Cloud?

You can deploy SAP HANA on the cloud for fast time-to-value or on-premise for maximum control. Both options have their pros and cons. The on-premise edition is an internal platform, located on your local servers. In this case, your local IT team has to implement periodical upgrades and run necessary tests.

SAP HANA Cloud edition is Software as a Service (SaaS) hosted on SAP’s servers and maintained by SAP. You can use the cloud edition to reduce your costs by replacing expensive hardware with pay-per-use services. This service offers more storage, computing, and networking capacity compared to on-premise systems. In addition, you get features like scaling and highly-available data centers across the globe. The cloud option also provides access to advanced infrastructure without the need to do repetitive maintenance.

SAP HANA on AWS

AWS provides a set of infrastructure services that you can use to deploy SAP HANA in a fault-tolerant, and highly available manner. This deployment enables you to leverage the flexibility and security of AWS together with the functionality of SAP HANA.

In addition, you can take advantage of the pay-per-use pricing model of AWS to pay only for what you use. You can choose between two licensing options for SAP HANA in AWS, either license per hour, or the Bring-Your-Own-License (BYOL) option.

Moreover, you can scale the SAP HANA infrastructure according to your needs and increase the level of availability in multiple Availability Zones (AZ). For storage, you can integrate SAP HANA with any EC2 instance and block storage option of your choice depending on your needs.

SAP HANA on AWS: Architecture

You can deploy the SAP HANA on AWS infrastructure either on a single node or multi-node architecture configuration.

Single-AZ, single-node architecture
The Single-AZ, single-node option provisions a single EC2 instance with the SAP HANA platform operating system. Make sure to configure a Virtual private cloud (VPC) with private and public subnets to secure your deployment. For secure access, you can place the SAP HANA server in the private subnet, which cannot be directly accessed from the internet. For secure SSH access to the SAP HANA server, you can use an SSH client or the bastion host on the optional Windows Server instance.

Multi-AZ, single-node architectureThe Multi-AZ, single-node option provisions two EC2 instances in two different AZs. This architecture is based on the SUSE Linux Enterprise Server (SLES) high availability extension. The extension is a part of the SLES for SAP operating system. There are two main SAP HANA high availability deployment scenarios:

• Cost-optimized — the secondary node can be used for production as well as non-production scenarios like QA. Before moving to production, you need to stop the non-production instance running on the secondary node.
• Performance-optimized — both secondary and primary nodes are of the same size. They utilize the synchronous replication mode for SAP HANA replication.

SAP HANA to AWS: Tips and Tricks

You can migrate SAP HANA to AWS by using the rehost or the replatforming migration approaches. Replatforming is useful for increasing system performance or adopting managed services. Rehosting means shifting your current systems to a cloud-based environment. Make sure to implement the following tips to make the most out of your migration strategy.

Understand on-premises resource utilization
If you are planning to migrate SAP HANA to AWS using the rehost migration approach, you need to understand the utilization of resources. This includes hardware configuration, network connections, and performance data, in your on-premises SAP HANA environment. This data can be used to enable communication channels between SAP HANA and other systems in VPCs or in AWS security groups.

AWS offers the Application Discovery Service to help you get an initial evaluation of how your on-premises environment is utilized, prior to migration.

Data tiering
If you are planning to migrate SAP HANA to AWS using the replatforming approach, you can distribute your data into cold and warm storage solutions like Hadoop on AWS or SAP HANA dynamic tiering. Right now, SAP supports HortonWorks, MapR, and Cloudera as Hadoop distributions for SAP HANA.

Determine memory requirements for rehosting
During rehosting, you can determine the size of the SAP HANA system you need on AWS based on the peak memory utilization of your on-premise system. Measuring peak memory utilization is a better method than measuring allocated memory because your on-premises SAP HANA environment may be oversized.

There are three methods to determine the maximum memory utilization of local SAP HANA systems:

• SAP EarlyWatch alerts — a free, automated service that enables you to monitor major administrative areas of your SAP system.
• SAP HANA Studio — provides a memory utilization summary in the overview tab in the administration view.
• SQL statements — you can use the SQL statements that SAP provides to determine peak memory utilization.

Determine the network bandwidth
You need to consider the required network bandwidth for transferring the amount of data to AWS. Network bandwidth impacts the data transfer time and the total downtime. Higher bandwidth leads to faster data transfer and reduces overall migration time. You can use smaller network bandwidth to reduce costs for non-production systems. In non-production systems, downtime isn’t critical.

SAP HANA instance sizing
AWS offers SAP-certified systems that are designed to meet SAP HANA performance requirements. After you have determined the sizing of SAP HANA, you can map the maximum amount of memory required for each of your SAP HANA instances to the maximum amount of available memory for EC2 instance type. You should also consider suitable storage volume sizes and types to ensure the optimal performance of the SAP HANA database.

Conclusion

The high performance of SAP HANA is both a benefit and an obstacle. The benefit is obvious, the issue with high performance is the expensive hardware and maintenance you have to purchase. Therefore, SAP partnered with Amazon to deploy SAP HANA in the cloud. This deployment enables you to take advantage of the benefits of SAP HANA and the cloud.