21 Fractal Analysis and Programming of Elastic Systems …
309
The chapter is organized as follows. In Sect. 21.2, the box-counting analysis of
the capacity curve of the elastic system is considered. In Sect. 21.3, the definition of
CCM is given. The model is integrated with the box-counting method. The fractal
control of the elastic object is introduced. In Sect. 21.4, the container-component
fractal analysis of the distributed gradient ascent algorithm is given. Section 21.5
concludes the chapter.
21.2 The Box-Counting Fractal Analysis of the Capacity
Curve
The core of the elastic computing is the workload and capacity. The assumption
is made that workload on the system equals to capacity of the system. The fractal
analysis base on box counting of the capacity curve of the elastic computer system
is represented in this section.
Hereinafter, the capacity curve of the elastic object is discussed in Sect. 21.2.1.
The analysis of the capacity curve is given in Sect. 21.2.2.
21.2.1 The Capacity Curve of the Elastic Object
The elastic concepts characterize a service. The incremental, decremental, and iterative nature of elastic object directly impacts of it construction and object hierarchies
in the design time of a complex software system. Due to this, the capacity of the
services needs to be incremented or decremented by IT resources with characterized
curve. Let workload on the system equals its capacity.
Fig. 21.1 Fluctuating demand of the users during the day is increasing and decreasing
309
The chapter is organized as follows. In Sect. 21.2, the box-counting analysis of
the capacity curve of the elastic system is considered. In Sect. 21.3, the definition of
CCM is given. The model is integrated with the box-counting method. The fractal
control of the elastic object is introduced. In Sect. 21.4, the container-component
fractal analysis of the distributed gradient ascent algorithm is given. Section 21.5
concludes the chapter.
21.2 The Box-Counting Fractal Analysis of the Capacity
Curve
The core of the elastic computing is the workload and capacity. The assumption
is made that workload on the system equals to capacity of the system. The fractal
analysis base on box counting of the capacity curve of the elastic computer system
is represented in this section.
Hereinafter, the capacity curve of the elastic object is discussed in Sect. 21.2.1.
The analysis of the capacity curve is given in Sect. 21.2.2.
21.2.1 The Capacity Curve of the Elastic Object
The elastic concepts characterize a service. The incremental, decremental, and iterative nature of elastic object directly impacts of it construction and object hierarchies
in the design time of a complex software system. Due to this, the capacity of the
services needs to be incremented or decremented by IT resources with characterized
curve. Let workload on the system equals its capacity.
Fig. 21.1 Fluctuating demand of the users during the day is increasing and decreasing