Configuration Management Tools, Statistical Analytics: A Review

This report provides a comprehensive overview of configuration management tools and statistical analytics. It begins by defining configuration management, emphasizing its role in tracking and maintaining hardware and software assets. The report explores features such as infrastructure automation, application deployment, and task management. It then delves into various configuration management tools like Microsoft SCCM, Ansible, Chef, and Terraform, detailing their functionalities. The report further explores statistical analytics, its benefits, and the different types of data involved, including transaction, personal, mobile app, and scientific data. It also differentiates between quantitative and qualitative data, discrete and continuous data, and nominal, ordinal, and binomial data types. Finally, the report addresses key challenges in data mining, such as scalability, high dimensionality, heterogeneous data, privacy and security issues, and performance issues, concluding with the importance of domain knowledge. The report includes a comprehensive list of references.

CHAPTER 2
LITERATURE REVIEW
Features And Capabilities Of The Configuration Management Tools
Configuration Management assures that all hardware and software assets that a company
owns are tracked and maintained at any instance. In simple terms, it’s astate-of-the-art
inventory for the assets maintained by the management. Configuration Management can be
achieved in an environment by consolidating two fields: software pipelines and
infrastructure-as-code. Earlier deals with technology to build and test the software artifacts
and later is the management practice of provisioning the infrastructure entities through
machine readable definition files.
Various configuration management tools for managing configurations e.g., installations of
application environments exist. These tools tend to be directed at specific vendor applications
and may be specific to certain hardware and software platforms. Administrators tend to be
wary of agents and daemons that run on their systems to perform various tasks for the most
part outside of their control or knowledge. In addition, agents and daemons typically
communicate with applications running remotely on some server or servers over the network.
Configuration management tools perform various roles to ensure consistency among physical
and logical assets. These tools identify and track configuration items and document
functional dependencies. They are invaluable for understanding the impact of changing from
one configuration item to others.
Features and capabilities of the configuration management tools are listed below
 Infrastructure automation
 Automated provisioning
 Application deployment
 Task management
 Visualization and reporting
 Orchestration
 Node management

 Role-based access control
The purpose of configuration management is to keep a detailed record of the information
about the computer system and to update them as needed. This includes listing all the
installed software, the network addresses of the computers, and the configuration of different
pieces of hardware, and creating updates or ideal models that can be used to quickly update
computers or restore them to a predefined baseline.
The major benefits are:
 Reduced risk of outages and security breaches
 Cost reduction by avoiding duplication of technology assets.
 Process control through enforcing formal policies and procedures Faster problem
resolution
 Efficient change management by understanding the baseline configuration
 Quicker restoration of service
2.1 Configuration Management Tools
Following are the list of Configuration Management tools and their details
S.No TOOL DESCRIPTION
1 Microsoft SCCM is system management
software for managing large groups of
workstations running Windows or any other
Operating Systems. It is one of the top-rated
configuration management tools.

2 Ansible is one of the leading configuration
management products from Redhat and can run
on both Unix-like and Windows OS. With the
complexities uprising due to virtualization and
cloud technologies, Ansible is designed is to
provide a consistent, reliable and secure solution
to manage the environment.
3 Bamboo is a continuous integration tool
developed by Atlassian. The main purpose of
Bamboo is used to automate the release
management, creating a continuous delivery
pipeline.
4 Chef is both the company and name of the
configuration management tool. As the managed
environments grow, manual configuration and
deployment practices compromises the
operational costs. To overcome the same, Chef
ensures that configuration policy is
flexible,testable, readable and can be versioned.
5 Terraform is a tool for building and versioning
infrastructure created by HasiCorp. With the
continual changes in the configuration,
Terraform can determine the changes and
creates incremental execution plans.
6 TeamCity is a build management and
continuous integration tool from JetBrains.
Some of the other features supported by
TeamCity are technology awareness, cloud
integrations, Code Quality tracking.
7 HashiCorp provides easy workflow for
developer, operator, and designer. It leverages a

declarative configuration file which describes all
your software requirements, packages, operating
system configuration, users, and more.
8 Puppet is a software configuration management
and deployment tool and mostly used on Linux
and Windows. It helps to pull the strings on
multiple application servers at once. We can use
Puppet on several platforms such as IBM
mainframes, Cisco switches, and Mac OS
servers
9 Salt is infrastructure management tool built on a
dynamic communication bus. It can be used for
data-driven orchestration, remote execution for
any infrastructure and configuration
management for any application.
2.2 STATISTICAL ANALYTICS
Statistical Analytics is the science of collecting, exploring and presenting huge amounts of
data and to infer significant information from the underlying dataset. Statistical Analytics is
closely associated with data mining domain, which is also functionally used to explore and
analyze the data to derive meaning patterns. Analytics is a part of everyone’s daily life and
are applicable in almost all disciplinary fields. Analytics provides realistic opportunities and
answer the two most important challenges in the data: automated prediction of trends and
behavior and automated discovery of previously unknown patterns. Some of the major
benefits that can be obtained from analytics are precise decision making, well-defined
planning and forecasting, novel revenue streams, competitive advantage and reinforced
customer relationships.
2.2.1 VARIETIES OF DATA
In an information era and in the age of data science, data generated from different disciplines
are unique in their own ways. The proper knowledge on them will make our approach to

VARIETIES OF DATA
Transaction data
Personal data Mobile app data
Surveillance data
Scientific data
analytics problem smarter. Here is the comprehensive detail of the various sources of data
and their nature.
Transaction data relates to the transactions of the organization and includes data that is
captured when the product is sold or purchased. A transaction is the sequence of information
exchange and it can be purchase order, shipping status, employee status, insurance and
medical claims etc
Scientific data is defined as the information collected using experimental methods for a
specific purpose of study and/or research. Example: data collected from the traffic generated
in the computer networks is a scientific data
Personal data deals with the identity of the person with the government or society. In
general, personal data contains entities such as name, identification number and location.
Personal identification data can be provided by government, society, organization or any
management.
Surveillance data helps to monitor the behavior, activities or information for the purpose of
information gathering or prevention of criminal activities. Some of the major sources of
Surveillance data are CCTV camera, Internet traffic, biometric sensors , etc.

Data Types
QualitativeQuantitative
Discrete
Continuous
Nominal
Ordinal
Binomial
Mobile app data refers to the data emitted from the smartphones such as user behavior,
general usage, interactions, cache files etc. Mobile analytics involves measuring and
analyzing data generated from mobile apps and some of the tasks achieved using the same are
mobile security, performance analytics and marketing analytics.
2.2.2 DATA TYPES
Analytics and data mining arethe mathematical sciences dealing with the collection, analysis
and interpretation of data. Understanding data types is predominant because different
statistical methods needs to be applied based on the nature of data. Hence proper
understanding on the various types of data will helps to solve the statistical analysis
problems.
The hierarchy above depicts the various types of statistical data. In a top level, two types of
data exist: Quantitative and Qualitative.
Quantitative data measures the data in the form of counts or numbers and each data attribute
has a unique numerical value associated with it. Quantitative data can always be verified and
can be evaluated using the statistical techniques. Some of the examples of Quantitative data
are dimensions: height, weight and length, weather: temperature, humidity and wind speed
and cost of the item in monetary units.

Discrete data is based on the counts and can have only finite number of values. The data
here is disconnected, distinct and unique. Examples of discrete data are number of students in
the class, and the number of customers who bought particular items.
Continuous data on the other hand is a measurable data and are taken over specific time
interval. Variables in continuous data will be having decimal points and they denotes precise
and accuracy of measures taken from the experiments. Examples of continuous data are
weight of newborn babies, daily wind speed and temperature of freezer.
Qualitative data can be observed and recorded and are often defined as data that
characterizes or approximates the data. Mostly, Qualitative data is non-numerical in nature
obtained from surveys, experimental observations and interviews. The significance of
Qualitative data can be observed as it allows the statisticians and researchers to derive
parameters from the Qualitative data. Data collection technique for Qualitative data is
exploratory in nature as it involves in-depth analysis and research.
Nominal data is used for naming or labeling variables without any quantitative value. There
is no intrinsic ordering followed for nominal data. Hence if the order of values is changed, the
meaning will not change. Nominal data can be analyzed using the grouping method. Nominal
variables can be grouped together into categories, and for each category, frequency or
percentage can be calculated.
Ordinal data on the other hand is natural and ordered categories and the distance between
categories is not known. Ordinal scale is distinguished from the nominal scale by means of
ranking. Example of Ordinal data can be found in survey questionnaires, Rate the quality of
solution provided by the telecom associate: Very good: 5, Good: 4, Average: 3, Bad: 2 and
Very bad: 1
Binomial data is a categorical data and have two values: True and False, Yes and No and
success and failure. Most of the real-time applications are of binary in nature and in such
cases binomial data are handy. For example: Building a supervised learning model to predict
whether tumor is benign or malignant.
2.2.3 CHALLENGES IN DATA MINING

Some of the major data mining issues are addressed below
Scalability: In real-world applications, dataset with sizes of terabytes and gigabytes are very
prevalent. Hence our analytics and data mining approach have to be well-equipped to handle
massive volume of data.
High dimensional: Nowadays, dataset holds thousands of attributes instead of countable
attributes few years ago. Complexity arises for the high dimensional data and there is a
demand to build efficient analytics model.
Heterogeneous and complex data: Data mining and analytics in their beginning stage
witnessed homogeneous data mostly of type - continuous or categorical. Currently,
applications of data mining are almost in all domains. Hence there is a necessity to grip on
heterogeneous data such as text, images, surveillance of audio and video, computer network,
clinical and biological etc.
Privacy and security issues: Most of the datasets are collected with the internet as the
medium. The vulnerabilities in internet may act on collected and generated data. Also, all
disciplines contain certain personal and confidential information about the specific person,
group or management. Data mining approach will be in a stage to take care of both the above
issues.
Performance issues: Most of the machine learning algorithms is designed with the
consideration of substantial size of data. But with the growing nature of high-dimensional,
heterogeneous and immense scalability of data, there is an obvious query on the adaptability
of algorithms on such scenario. This will either drain the computational resources or may
demand new approach to handle the complexities. In general, performance issues in data
mining depend on the statistical algorithms, platforms used, and the nature of tools and
software involved.
Domain knowledge: If domain knowledge can be incorporated, more reliable and accurate
data mining solutions can be found. Both descriptive and predictive tasks can come up with
more useful findings and can make more accurate predictions. But collecting and
incorporating domain knowledge is a complex process.

REFERENCES:
1. Scratching the Surface of Windows Server 2016 and System Center Configuration Manager
Current Branch
2. Mastering System center Configuration Manager
3. Proposal and Implementation for Highly Available Solution of Virtual Infrastructure and VDI
4. Configuration Manager - an overview, ScienceDirect Topics
5. Microsoft System Center 2012 Configuration Manager: Administration Cookbook
6. Microsoft System Center Configuration Manager Advanced Deployment
7. https://www.interlink.com/solutions/microsoft-azure/microsoft-system-center-configuration-
manager
8. https://www.prajwaldesai.com/sccm-2012-r2-step-by-step-guide/
9. https://docs.microsoft.com/en-us/mem/configmgr/develop/core/understand/sqlviews/create-
custom-reports-using-sql-server-views
10. https://docs.microsoft.com/en-us/mem/configmgr/desktop-analytics/overview
11. https://www.anoopcnair.com/sccm-desktop-analytics-integration/
12. https://www.systemcenterdudes.com/sccm-desktop-analytics/
13. Data analytics and management of computing infrastructures
(https://patentimages.storage.googleapis.com/f2/9a/a5/167a6cd5fbe483/
US20170034016A1.pdf)
14. https://www.professionalqa.com/software-configuration-management-tools
15. https://www.softwaretestinghelp.com/top-5-software-configuration-management-tools/
#:~:text=In%20Software%20Engineering%20Software%20Configuration,in%20the
%20establishment%20of%20baselines.
16. https://docs.microsoft.com/en-us/mem/configmgr/core/understand/introduction
17. https://www.prajwaldesai.com/setup-sccm-desktop-analytics-connect-sccm-with-desktop-
analytics/
18. https://blog.juriba.com/windows-10-assessment-upgrade-readiness#:~:text=Microsoft
%20Upgrade%20Readiness%20was%20released,on%2Dpremise%20and%20cloud
%20environments.
19. https://www.prajwaldesai.com/sccm-2012-compliance-settings/
20. https://mindmajix.com/sccm-tutorial
21. http://habib-sccm.blogspot.com/p/version-history.html

22. https://www.trustradius.com/configuration-management
23. https://www.ansible.com/use-cases/configuration-management
24. https://confluence.atlassian.com/bamboo/understanding-the-bamboo-ci-server-
289277285.html#:~:text=Bamboo%20is%20a%20continuous%20integration,creating%20a
%20continuous%20delivery%20pipeline.
25. https://www.terraform.io/intro/index.html
26. https://www.vagrantup.com/