In the field of computing theory or computer science in general, there are many problems that are computationally hard. These problems attract many researchers who try to solve them. But one very important aspect needs to be done before solving these problems: they need to be classified, or defined, well, so that people can work effectively and efficiently to minimize redundant efforts. This paper attempts to contribute to this aspect by presenting a general taxonomy for scheduling problems and solutions in distributed systems.

Scheduling problems have been recognized as being computationally hard for decades. They include a huge pool of problems, without good classifications. This paper investigates a complete list of state-of-the-art research products for solving these problems and recognizes 109 such problems. Furthermore, the problems are grouped into ten sections so that each group employs similar features. The taxonomy can be applied to any kind of distributed platform, including clusters, cloud computing, and grids.

The contribution of the paper is very appreciated since it can be used as a “problem pool” for researchers in the distributed system field, and it can also be used as a guidebook or checklist for the community to record what has been done. Overall, I hope there are similar efforts in other fields that include similar hard problems without good classification systems.