SCP - Science Cloud Platform PDF Print E-mail

The autonomic cloud case study is realized in a prototype implementation called the “Science Cloud Platform” (SCP) - a distributed software system which is able to execute applications in the presence of challenges such as

  • A fluctuating environment: heterogeneous nodes which join and leave at will
  • Different requirements of applications such as required CPU time and memory that must be satisfied
  • The need for energy conservation wherever possible, but still being able to execute applications

A working demonstrator of the ASCENS cloud has been implemented which allows:

  • It allows executing applications “on the cloud” in a heterogeneous, p2p-based network.
  • It reacts to events such as leaving and joining nodes, or increased load
  • It can use the Zimory IaaS platform for dynamically adding and removing virtual machines for the network

Each application in the cloud is managed by several nodes in different roles

  • A node with the initiator role is responsible for supervising app storage, migration, and execution
  • A node with the executor role is responsible for actually executing the app
  • Different other roles support the management process by storing app bytecode and deploying / undeploying apps and virtual machines

Further Information:

Last Updated on Monday, 09 March 2015 15:09