If you are facing performance issues and would like to extend your server-side hardware, a draft of a reliable, scalable solution has been sketched with high uptime and maintainability based on Microsoft technologies.
It requires two machines (further Nodes) with installed Windows Server 2008 (or Windows Server 2003 as minimum).
The nodes have copies of Microsoft services installed:
– Microsoft Cluster Server, which makes the cluster
– MS SQL, which is a database
– SSRS, which is a report server
– SSAS, which is an analysis server
– IIS, which is a web server
The nodes and services form a single virtual unit called Cluster which provides a unified interface to Internet/Intranet.
The nodes sync between each other using Private Cluster Network (further Heart Beat)
Node #1 runs MS SQL, SSRS and IIS in operation mode, while Node #2 runs SSAS.
If a failure happens or a maintenance is required to one of the nodes or services, the other node runs up missing services automatically.
All data is stored in Storage Area Network (further SAN), which is RAID. Optional, the SAN could be mirrored with a separate node which is running MS SQL, for more reliability.
Optional, the SAN could be accessed via SAN Switch as a mapping solution.
Likely, if you have ready applications, applications’ infrastructure will not be affected by the solution, and code base changes are not required.
See the draft below: