CactoSim Version 2 Released

0 Flares 0 Flares ×

Cloud computing data centres are complex systems with a high degree of heterogeneity and a large number of different elements (e.g. racks, nodes, hard disk drives, virtual machines) with various forms of interactions and dependencies. Any system with these characteristics is exceptionally difficult to manage as any decision to make a change or react to an exception can have significant operational and cost implications.

placeit3-300x225 To support decisions of this nature, the inherent relationships within the system must be understood and incorporated into an evaluation process. The use of CactoSim gives the ability to model system behaviour and produce information for large-scale data centres without starting a single Virtual Machine.

CactoSim is currently released as Version 2, the “CactoSim Intermediate Prototype” release. This release provides closer integration between simulation and the real data centre components, as part of the CACTOS project. The focus falls upon deeper integration with the CactoScale monitoring framework and additional new features of integrating with the CACTOS Runtime Toolkit.

CactoSim plays the role of a context-aware decision-support tool for data centre management. Data centre operators are able to retrieve up-to-date automatically created models of heterogeneous hardware and virtual landscape components from the Runtime Model Storage provided by CactoScale.

download-cactosim-buttonBy using CactoSim, these obtained real system models then can adapted further in order to validate and evaluate optimisation algorithms provided by CactoOpt via what-if-analyses. Produced results, in the form of simulated predictions, are derived from collected historical system data such as resource utilisation, usage patterns, energy consumption and existing optimisation strategies.

The CactoSim Intermediate Prototype is made available under the Eclipse Public License, EPL Version 1.0.

 

 

0 Flares Twitter 0 Facebook 0 Google+ 0 LinkedIn 0 Email -- 0 Flares ×
The following two tabs change content below.
James is a Senior Researcher in the Dublin City University Business School. Prior to this he worked as a Specialist Researcher for SAP Research. His research interests include Web and Cloud based simulation, supply chain management, and discrete event simulation. In the past he has collaborated with a number of academic and industrial partners. He holds a B.Tech in Information Technology and Telecommunications, an M.Tech in Computer Integrated Manufacturing and a Ph.D from the University of Limerick. His Ph.D is in the area of Web based simulation applied to contract costing for outsourcing enterprises.