Class RL2Profile

All Implemented Interfaces:
Serializable, LoadBalancingProfile

public class RL2Profile extends AbstractLoadBalancingProfile
Parameters profile for the "rl2" algorithm.
Author:
Laurent Cohen
See Also:
  • Constructor Details

    • RL2Profile

      public RL2Profile(TypedProperties config)
      Initialize this profile with values read from the specified configuration.
      Parameters:
      config - contains a mapping of the profile parameters to their value.
  • Method Details

    • getPerformanceVariationThreshold

      public double getPerformanceVariationThreshold()
      Get the variation of the mean execution time that triggers a reset of the bundler.
      Returns:
      the variation threshold as a double value..
    • getPerformanceCacheSize

      public int getPerformanceCacheSize()
      Get the maximum size of the performance samples cache.
      Returns:
      the cache size as an int.
    • getMinSamples

      public int getMinSamples()
      Get the number of known states from which to start computing a probablity that the next action will be chsen randomly.
      Returns:
      the min samples size as an int.
    • getMaxSamples

      public int getMaxSamples()
      Get the number of known states from which the learning phase is considered complete.
      Returns:
      the min samples size as an int.
    • getMaxRelativeSize

      public double getMaxRelativeSize()
      Get the maximum bundle size, expressed as a fraction in the range ]0.0, 1.0], of the current job size.
      Returns:
      the max relative size as a double value between 0 and 1.
    • toString

      public String toString()
      Overrides:
      toString in class Object