Server farms are ubiquitous in applications ranging from Web server
farms to high-performance supercomputing systems to call centers. The
popularity of the server farm architecture is understandable, as it
allows for increased performance, while being cost-effective and
easily scalable, allowing applications to meet response time goals and
power goals.

Given the prevalence of server farms, it is surprising that even at
this late date so little is understood with respect to designing
policies for resource management in multi-server systems.  In this
talk we address the following design questions:

** What are good policies for routing/dispatching jobs to hosts in a server farm?  
** How should jobs be scheduled at the individual hosts?
** How can we choose policies to combat job size variability?
** What are the tradeoffs between response time and fairness goals?
** How can policies simultaneously meet response time goals and 
   energy goals for a server farm?

There are at least three distinct communities studying scheduling in
server farms from an analytical perspective.  These include the
SIGMETRICS community, the INFORMS community, and the SPAA/STOC/FOCS
community, all of which have different approaches and goals.  One of
our goals in this tutorial is to make researchers aware of results in
these different communities.  The emphasis will be on intuition, so
that the talk is accessible to newcomers as well as old-timers. In
surveying the newest results, we will also present some practical open
problems.