I want to know that what are the different research areas left in the problem of optimal task scheduling in distributed computing. I have refereed many papers for this but exactly not able to find the gaps which are left. Please answer this with reference.
Whether precedence-constrained scheduling of tasks on 3 (or more processors) is NP-complete or not is still an open issue, I think (cf. http://dl.acm.org/citation.cfm?doid=1077464.1077476). This is irrespective of if they are truly distributed or "merely" parallel.
If task scheduling is available then not all data structures are distributed, and your system is not truly distributed, it is a hybrid. Sorry I cannot help.
If task scheduling is available then not all data structures are distributed, and your system is not truly distributed, it is a hybrid. Sorry I cannot help.
However, totally amorphous systems [random architecture, random activity, local communication only] can recognize simple graphs which are tripartite [NP-complete in Turing computability] by producing a partition. Also, looking back at my earlier answer, I probably should have asked for a precise definition of the scheduling problem you had in mind.
- Badges
- Science method