Multiprocessor Real-Time Scheduling

1
Multiprocessor Real-Time Scheduling

• Aaron Harris
• CSE 666
• Prof. Ganesan

2
Multiprocessor Scheduling

• As more systems incorporate multiple processors,
  interest in scheduling algorithms increases.
• Still relatively little research done (compared
  to uniprocessor scheduling).
• Due to increased complexity, much research makes
  simplifications of their systems.

3
When and Where

• Multiprocessor scheduling requires determining
  both which task to run at a given time and on
  which processor to run it on.
• Added complexity as compared to scheduling in a
  uniprocessor environment.
• Scheduling must account for not only the status
  of tasks, but the status of processors.

4
Schedule Management

• In a standard system, the schedule is controlled
  by the processor.
• Multiprocessor systems follow two general
  approaches master-slave and dedicated scheduler.
• In master-slave, one processor handles all
  schedule processing and also processes normal
  tasks.
• Dedicated scheduler systems allocate an entire
  processor to nothing but schedule processing.
  This can be more efficient in complex systems
  with many processors.

5
Algorithm Selection

• In a single processor system, rate-monotonic (for
  static priorities) and earliest-deadline-first
  (for dynamic priorities) algorithms are the
  accepted best overall algorithms.
• There is no agreed-upon best algorithm for
  multiprocessor systems (due to increased
  complexity of algorithms and lack of research).

6
Algorithm Categories

• Most multiprocessor scheduling algorithms fall
  into one of two categories global scheduling and
  partitioned scheduling.

7
Partitioned Scheduling

• All tasks are divided among the processors at
  compile time.
• Each task can only execute on its assigned
  processor.
• Advantages This turns each processor into a
  single processor system as far as scheduling is
  concerned. All research and theory of single
  processor scheduling then applies to each
  processor.
• Disadvantages The process of assigning tasks to
  processors can be complicated and extremely
  inefficient if done poorly. This can lead to
  poor system performance.

8
Global Scheduling

• All tasks are allowed to execute on all
  processors.
• Tasks are scheduled from a global queue to
  individual processors.
• Advantages This technique allows for better
  processor utilization and overall system
  performance.
• Disadvantages The main disadvantage is
  migration. This is when a task is preempted on
  one processor and resumes execution on a
  different processor. This requires the context
  of the task to be moved from one processor to
  another, which can be costly.

9
Problems Dhalls Effect

• Something interesting is the relation of
  ultization to scheduliblity.
• One may think that to be schedulable, the
  utilization of a set of tasks on n processors
  should simply be n.
• Dhall proved that the number must actually be
  between nln(2) and n (Dhalls Effect).

10
Problems Schedule Complexity

• Many types of scheduling (e.g. dynamic
  priorities) that require extensive run-time
  calculations can become too complicated in a
  multiprocessor environment.
• Due to this, most of the scheduling algorithms
  that have been researched use compile-time
  determined priorities.

11
Problems Resource Synchronization

• Resource synchronization becomes a much larger
  issue in multiprocessor systems.
• Even in a non-preemptive multiprocessor system,
  resource sharing can lead to task blocking.
• More thought must be put into how and what
  resources are shared between tasks in a
  multiprocessor system.

12
Conclusions

• Multiprocessor scheduling algorithms are more
  complex than uniprocessor systems.
• There is still no decisively best multiprocessor
  scheduling algorithm.
• Multiprocessor systems will continue to become
  more and more prevalent, and as a result,
  multiprocessor scheduling research and theory
  will have to continue to improve.

13
References

• Khemka A., Shyamasundar R. K., "An Optimal
  Multiprocessor Real-Time Scheduling Algorithm",
  Journal of Parallel and Distributed Computing 43,
  37-45, 1997 http//citeseer.ist.psu.edu/khemka97op
  timal.html.
• G. Manimaran, C. Siva Ram Murthy, "An Efficient
  Dynamic Scheduling Algorithm for Multiprocessor
  Real-Time Systems," IEEE Transactions on Parallel
  and Distributed Systems, Vol. 09, No. 3,  pp.
  312-319, March, 1998.
• K. Ramamritham, J.A. Stankovic, P.F. Shiah,
  "Efficient Scheduling Algorithms for Real-Time
  Multiprocessor Systems," IEEE Transactions on
  Parallel and Distributed Systems,
  Vol. 01, No. 2,  pp. 184-194, April, 1990.
• J. Carpenter et al., A Categorization of
  Real-Time Multiprocessor Scheduling Problems and
  Algorithms, http//www.cs.uga.edu/shelby/pubs/ca
  rpenterFHSAB2003.pdf.
• S. K. Dhall, C. L. Liu, On a Real-Time
  Scheduling Problem, Operations Research, Vol.
  26, No. 1, Scheduling (Jan. Feb., 1978), pp.
  127-140.

14
Questions?