Method and apparatus for performance and policy analysis in distributed computing systems

1. A method for analyzing one or more policies applicable to a plurality of computing sites that process respective workloads in a distributed computing system, said method comprising:

using a processor to perform steps comprising:

representing said distributed computing system as a state transition model, wherein said representing comprises:

constructing said state transition model in accordance with at least one model representing characteristics of said plurality of computing sites, where said at least one model includes at least one of: a site model, a workload model, a workload state model, a site state model, a policy model, an event model, or a cost model, wherein said constructing comprises:

modeling, at a first computing site of said plurality of computing sites, an availability of associated resources as a state;

modeling, at said first computing site, changes to an associated workload as a first set of state transitions and changes to said availability of associated resources as a second set of state transitions; and

modeling one or more policies that dictate a manner in which said first computing site shares resources with a second computing site of said plurality of computing sites as constraints on said state and on said first set of state transitions and said second set of state transitions;

superimposing a queuing network model on said state transition model, wherein said superimposing said queuing network comprises:

associating each state transition in said first set of state transitions with a first probability that indicates a likelihood that said each state transition in said first set of state transitions will occur, where said first probability depends upon one or more characteristics of said associated workload;

associating each state transition in said second set of state transitions with a second probability that indicates a likelihood that said each state transition in said second set of state transitions will occur, where said second probability depends upon one or more characteristics of failure and recovery of said associated resources;

associating said first probability with a first policy of said one or more policies and said second probability with a second policy of said one or more policies;

associating each of said first policy and said second policy with a third probability that indicates a likelihood of at least one action associated with the first policy or the second policy being taken; and

deriving at least one fourth probability comprising a steady state probability of at least one state of said distributed computing system in accordance with a queuing network analysis technique; and

identifying one or more effects of said one or more policies on said distributed computing system performance in accordance with a solution to said queuing network model.