Microsoft Failover Cluster with Shared Hosts

Placing one data tier and one SAP Mobile Server node, installed as an application server, on each of two physical servers, is a modification of the standard Microsoft Failover Cluster. This is the minimum hardware configuration that ensures there is no single point of failure in the system.

Microsoft Failover Cluster on Shared Hosts

Summary of Microsoft Failover Cluster architecture on shared hosts:

Resembles two single-server installations, each with an SAP Mobile Server node and a data tier installed on a single server. However, the data tiers are also incorporated in a Microsoft Failover Cluster, so that if the active data tier node fails, the passive data tier node becomes active and database support continues uninterrupted.
Consists of two data tiers installed in a Microsoft Failover Cluster support the SAP Mobile Server nodes. If the active data tier node fails, the passive data tier node becomes active and database support continues uninterrupted.
A proxy server farm can optionally provide enhanced load balancing for distributing requests to the two SAP Mobile Server nodes, but if you select the shared host configuration because of cost considerations, you may also want to omit proxy servers.
A hardware load balancer can optimizes the performance of a proxy server farm, if one is present.

Design Characteristics and Use Cases
Consideration	Details
Recommended for	Multiuser development and test environments of larger scale, where economizing on hardware costs is of prime importance. Moderate-volume production environments that require some high availability.
Degree of collocation	Maximum collocation possible without a single point of failure.
Not recommended for	High-volume production environments.
Worksheet to complete	Install MS Failover Cluster worksheet. See Completing Installation Worksheets.
Installation procedure used	Installing SAP Mobile Server with a Microsoft Failover Cluster with Shared Hosts in Installation Guide for Runtime.
Limitations	Host systems must be adequate to support collocation (if used). Disk resources must be adequate to support all databases managed by the data tier servers. Increased complexity of managing and monitoring the server hardware and its associated software. Higher infrastructure costs. Increased response times due to the increased load on the secondary server or the need to synchronize state information on or from multiple servers.
Benefits	Improved performance. No single point of failure. Easily accommodates routine maintenance and upgrading. Because data tier is in a failover cluster, this design allows for system downtime without affecting availability.