Home > Workload Solutions > SAP > Guides > VMware Virtualized SAP HANA with Dell Storage > SAP HANA deployment options
As in physical environments, you can install an SAP HANA system in virtualized environments as a scale-up single-node system or as a scale-out multinode system. The number of SAP HANA virtual machines that you can install on a single physical server depends on a number of factors, including the number of CPU sockets, the server RAM, the SAP HANA installation use case (production or non-production), and the maximum scalability that SAP and VMware support.
When you create an SAP HANA virtual machine and need to size and configure memory, CPU, hyperthreading, and non-uniform memory access (NUMA) nodes for the virtual machine, see the relevant VMware best practices and recommendations.
Scale-up systems are primarily used for transactional SAP applications such as the SAP Business Suite. The SAP HANA database is stored in the RAM of a single virtual machine, while the SAP HANA persistence is stored on a data and log file system belonging to this virtual machine.
Figure 1 shows the virtual disk devices used in SAP HANA virtual machines. The virtual disk that holds the operating system is created during the initial machine configuration and can reside on any available vSphere datastore. Installing the SAP HANA software requires a /hana/shared mount point. In scale-up installations, this mount point is private to the virtual machine and can be a directory under the operating system root device (if the size of the operating system root device meets the operating system and /hana/shared capacity requirements) or a separate virtual disk mounted under /hana/shared. This virtual disk does not have specific performance requirements and can reside on any multipurpose datastore.
Scale-out systems are primarily used for analytic applications such as SAP Business Warehouse. The SAP HANA database is distributed across multiple virtual machines, either on a single physical server or across a vSphere cluster consisting of multiple physical servers. The system uses a “shared-nothing cluster architecture,” in which every node in a scale-out deployment stores its database persistence on its own data and log file systems. In a scale-out deployment, the /hana/shared file system must be shared across all SAP HANA nodes that are part of the scale-out cluster. For configuration guidelines, see SAP HANA shared file system.