Open Access asks:
I see photos showing multiple racks of servers. How are they functioning together? How many processor cores are allocated to a website? I’ve read that there’s a limit to the processing speed of a single processor. What’s the hardware architecture which is providing the power to handle large amounts of traffic?
2 Answers -
Bruce Epper
June 27, 2012Just because servers are racked together does not mean they all are serving the same purpose, but when they are, they are generally configured as a server or application cluster. By setting up servers in a cluster, they provide fail-over functionality or load-sharing functionality. Fail-over will switch services in the event of a hardware or major software failure on one unit to another standby system. Load-sharing or load-balancing does exactly what is implied by the name. The same set of services and functions are running on all of the clustered machines and a manager process will allocate the next available machine by either a round-robin or least-used mechanism.
In a cluster setup, the hardware architecture used is generally irrelevant as a cluster can use pretty much any kind of hardware (you can find clustering software for most hardware out there). The key to handling large traffic volumes tends to be with optimization of services and good cluster management.
If you want all of the gory details of how it is done, search the web for server clusters. There are also application clusters which are similar but are dedicated to things like high-volume database applications such as those used by financial institutions and the NASDAQ among others.
Irshaad Abdool
July 11, 2012servers usually share workloads so as not to overload their processors.
this link may help http://www.howstuffworks.com/web-server.htm