Analysis of Distributed Information Systems: Scalability, Middleware

This essay provides a detailed overview of distributed information systems, defining them as networks of autonomous systems connected by middleware for resource sharing. It highlights key characteristics such as resource sharing, cost-effectiveness, and transparency. The essay delves into the role of middleware in simplifying communication and managing distributed applications, as well as the concept of scalability, discussing load, space, and structural scalability with examples. It also outlines various approaches to building distributed systems, including message-oriented and publish-subscribe middleware. Furthermore, the essay addresses the difficulties faced by management, such as inherent and accidental complexities, inadequate methods, and continuous re-discovery of techniques, concluding that distributed systems function as a single computer with high performance.

Running head: DISTRIBUTED INFORMATION SYSTEM
Distributed Information System
Name of the Student
Name of the University
Author’s Note:

1
DISTRIBUTED INFORMATION SYSTEM
A distributed system can be defined as the network, which comprises of the
autonomous systems connected by utilizing any distributed middleware [6]. This type of
systems is utilized to share the various resources as well as capabilities for providing its
users with an integrated and single network. The most important characteristics of this
distributed system mainly involve allowing sharing of resources like software and systems,
effective, efficient, cost effective and could be utilized from anywhere [3]. The main
objective of this distributed system is to provide transparency, reliability and openness to
the network and thus making it extremely popular for the users.
The following essay outlines a brief discussion on the concept of distributed
information system with relevant details. The essay will highlight the details of middleware
and scalability in distributed system. Moreover, the various approaches and methods for
building this distributed system will be mentioned here. The difficulties will also be provided
here.
1. Middleware and its Role in Distributed System
Middleware can be defined as the computer software, which gives services to the
software applications that are not available in the operating system. Middleware is
responsible for making the implementation of communication as well as input output
extremely easy by the developers of software [1]. This helps in focusing on the particular
motive for their application. It lies between the applications and the operating system on
both the sides of the distributed computing network.
Middleware plays an important role in the distributed system. It is the particular
type of infrastructure that helps to facilitate the creation of all applications of business and
thus providing major services such as transactions, messaging, concurrency, threading and
the framework for service component architecture for the SOA applications [5]. The
distributed nature of any application within the distributed system is hidden by this
middleware. Moreover, the interconnected parts are kept running as well as operational
within the distributed locations thus making these extremely simpler and easier in
managing.

2
DISTRIBUTED INFORMATION SYSTEM
2. Scalability and its Three Major Types
Scalability can be defined as the sole capability of any network, system and process
for handling the increasing amount of work or with the potential for accommodating the
growth [8]. Scalability is an extremely important requirement that makes any system
effective and efficient. Any system becomes scalable if it has the capability for increasing the
total output.
There are three major types of scalability in distributed systems. The descriptions of
each type with examples are given below:
i) Load Scalability: A system has load scalability when it comprises of the capability
in functioning gracefully. It occurs only when there is no delay in the functions and no
consumption of unproductive resources [2]. The load scalability can eventually schedule the
shared resources and helps in inadequate parallelism exploitation. The example of this type
of scalability refers to the system that has self expanding performance measure.
ii) Space Scalability: A system has space scalability when its memory requirement
does not develop to the intolerable level since the number of items is increasing [7]. The
example of space scalability is the sparse matrix method or compression.
iii) Structural Scalability: A system has structural scalability when the
implementation or standard does not impede the increment of object numbers [3]. The
example of structural scalability is any system with finite address space.
3. Various Approaches to build Distributed System
There are various approaches and methods for building any distributed system. They
are as follows:
i) Message Oriented Middleware: This type of approach helps to structure the
communication and thus the distributed system is built. The main benefit of this particular
approach is that it helps to support for the asynchronous communication [1]. Hence, the
sender transmits data to the receiver without waiting for any response.
ii) Publish or Subscribe Middleware: This type of approach is for the synchronous
communication. They are responsible for producing events on various topics, which are

3
DISTRIBUTED INFORMATION SYSTEM
propagated with the system [5]. The publishers require describing the kind of events
generated priori.
4. Difficulties faced by Management
There are various difficulties and technical issues that are being faced by the
management in any distributed system. The most significant issues faced by them are given
below:
i) Inherent Complexities: The inherent complexities mainly arise from the domain
challenges like elements of distributed system residing in different address spaces and
protocols utilized for communications [1]. The various networks, which connect the
components with the distributed systems, involve several problems like overload, transient
failures and latency that eventually affects the efficiency of the system.
ii) Accidental Complexities: This type of problem mainly arises from the limitations
with the software tools or techniques like poor distributed debuggers and non portable APIs
[4]. These types of problems are often deliberate by the developers, who like lower level
languages or platforms like C based operating system.
iii) Inadequate Methods or Techniques: The various methods of software analysis
and techniques for design cannot produce the best quality service or QoS [7]. This is
extremely dangerous and thus the developers face problems due to the inadequate
techniques.
iv) Continuous Re-discovery of Techniques: The re-invention of the techniques are
continuous and thus the problems are increased for the developers [6]. The real time
operating systems often become difficult to manage and control hardware resources.
Therefore, from the above discussion, it can be concluded that, any distributed
system is the specific model, where the components are located on the networked
computers or systems for communicating and coordinating the activities by simply passing
messages. These components eventually interact amongst each other for achieving the
similar goal. There are three important features of this distributed system and they are the
lacking of a global clock, components concurrency and sole failure of components. These
types of systems work and act as one single computer. The performance of these systems is

4
DISTRIBUTED INFORMATION SYSTEM
much higher than the other distributed models. The above essay has outlined a detailed
description on the concept of distributed system and how this system could be utilized by
the users. The middleware and its role in distributed system are also provided here.
Moreover, the scalability and the various approaches for building the system are mentioned
here. The essay has also given the difficulties faced by management due to distributed
system.

5
DISTRIBUTED INFORMATION SYSTEM
References
[1] Abadi, Martín, Ashish Agarwal, Paul Barham, Eugene Brevdo, Zhifeng Chen, Craig Citro,
Greg S. Corrado et al. "Tensorflow: Large-scale machine learning on heterogeneous
distributed systems." arXiv preprint arXiv:1603.04467 (2016).
[2] Chen, Tianqi, Mu Li, Yutian Li, Min Lin, Naiyan Wang, Minjie Wang, Tianjun Xiao, Bing Xu,
Chiyuan Zhang, and Zheng Zhang. "Mxnet: A flexible and efficient machine learning library
for heterogeneous distributed systems." arXiv preprint arXiv:1512.01274 (2015).
[3] Orgerie, Anne-Cecile, Marcos Dias de Assuncao, and Laurent Lefevre. "A survey on
techniques for improving the energy efficiency of large-scale distributed systems." ACM
Computing Surveys (CSUR) 46, no. 4 (2014): 47.
[4] Valls, Marisol García, Iago Rodríguez López, and Laura Fernández Villar. "iLAND: An
enhanced middleware for real-time reconfiguration of service oriented distributed real-time
systems." IEEE Transactions on Industrial Informatics 9, no. 1 (2013): 228-236.
[5] Ports, Dan RK, Jialin Li, Vincent Liu, Naveen Kr Sharma, and Arvind Krishnamurthy.
"Designing Distributed Systems Using Approximate Synchrony in Data Center Networks."
In NSDI, pp. 43-57. 2015.
[6] Enokido, Tomoya, and Makoto Takizawa. "An integrated power consumption model for
distributed systems." IEEE Transactions on Industrial Electronics 60, no. 2 (2013): 824-836.
[7] Stubbs, Joe, Walter Moreira, and Rion Dooley. "Distributed systems of microservices
using docker and serfnode." In Science Gateways (IWSG), 2015 7th International Workshop
on, pp. 34-39. IEEE, 2015.
[8] Ahmed, Waseem, and Yong Wei Wu. "A survey on reliability in distributed
systems." Journal of Computer and System Sciences 79, no. 8 (2013): 1243-1255.