Distributed computing is a computer science, which studies how a huge computing power needed to solve the problem into many small parts, and then assign these parts to many computer processing, the final results of these calculations together to get the final results.
Task scheduling in heterogeneous distributed computing using genetic algorithms Andrew J. Page and Thomas J. Naughton Department of Computer Science, National University of Ireland, Maynooth, Ireland. http://www.cs.may.ie/distributed 1 Outline Introduction Distributed Computing Scheduling algorithm Results Conclusion Future Work 2 Introduction Dynamic task allocation problem Heterogeneous distributed computing NP-Hard problem Evolutionary algorithms well suited Other solutions less efficient Zero communication time Created scheduler to more realistically model distributed systems 3 Distributed Computing Client-server model 4 Distributed Computing Variable heterogenous computing resources Non-exclusive usage Processors can be added or removed Network can become congested 5 Distributed system Utilises spare clock cyles from desktop PCs 200+ PCs in NUI Maynooth utilised Multiple operating systems Windows 98/NT/2000/XP, Linux Fedora/Debian/Mandrake, Apple, Solaris 6 Speedup - TSP 7 Scheduling ? ? ? ? ... Static Dynamic •Schedules created before •Schedules created during runtime & cannot change runtime •Tasks must all be known •No knowledge of task in advance until it arrives •Cannot adapt to changing •Can adapt to changing resources resources 8 Scheduling Assigned Scheduler Queue Immediate empty Batch ... 9 Scheduling Algorithm Genetic Algorithm based Search large spaces quickly List scheduling heuristic Batch scheduling Smoothing function Linpack benchmark 10 Scheduling Algorithm Initialisation: Most-into-least list scheduling heuristic Fitness function: based on relative error Euclidiean distance from lower bound to current solution Roulette wheel selection Cycle crossover Elitism 11 Experiments Micro GA Population size of 10 Poisson randomness Normal, uniform, and multimodal also tested Compared to 6 schedulers 3 immediate mode 3 batch mode 12 Other schedulers Immediate mode Earliest first Lightest Load Round Robin Batch mode Min-Min Max-Min Zomaya: homogeneous state-of the-art 13 Setup Simulation Lookup tables in scheduler Processor execution rates Communications costs Model communications & processors Queue of assigned tasks 14 Efficiency High % Comms Low % Comms 1/time spent communicating 15 Conclusion Dynamic task scheduler for heterogeneous distributed computing Statistical estimation of system properties Better performance than existing techniques 16 Future Work Deploy on live distributed system Generalise futher Use other evolutionary techniques Tabu, Ant Colony, etc... Employ better pattern recognition techniques 17 Contact Us Homepage: www.cs.may.ie/distributed Email: email@example.com Support is acknowledged from the Irish Research Council for Science, Engineering, and Technology, funded by the National Development Plan. 18
Pages to are hidden for
"Task scheduling in heterogeneous distributed computing using "Please download to view full document