Engineering, Technology & Applied Science Research Vol. 5, No. 5, 2015, 850-853 850

Engineering, Technology & Applied Science Research
Vol. 5, No. 5, 2015, 850-853
851 
www.etasr.com 
Erturk and Jyoti: Perspectives on a Big Data Application: What Database Engineers and IT Students… 
 
structured query language (SQL) statements of traditional 
relational databases. There are many NoSQL document 
oriented databases, for example, CouchDB and MongoDB. 
Document oriented databases store and retrieve data according 
to the meta-data definitions and tags found in their documents.
MongoDB was first developed by in 2009. It has versions 
that are compatible with different editions of Windows, Linux, 
Solaris, and Mac operating systems. MongoDB is an open 
source and free database application, under the GNU Affero 
General Public License. It is currently the most popular 
application in the category of document oriented databases [5].
MongoDB can store semi-structured and polymorphic data 
as well as structured data. Much of the current big data 
exchanged on the internet does not follow clear cut structures, 
rigid schemas, or restrictions in terms of data type and length. 
This also means handling emails, forums, complicated large 
objects, and multimedia files [6]. MongoDB allows users to 
flexibly define arrays within documents, and perform various 
operations on those array fields. Furthermore, MongoDB offers 
database querying functionality and supports high performance 
indexing. These are useful, for example, for text searches and 
manipulation of geospatial data. Instead of SQL Join 
statements, MongoDB users may reference a document from 
another or embed a document inside another document. It also 
features a powerful data aggregation and data analysis 
framework, for performing calculations on large data sets.
MongoDB aims to provide high performance, availability, 
and scalability, for cloud based information systems and big 
data environments. In order to ensure availability, sophisticated 
database systems create replicas of database files in different 
locations so that, if one location is down, then the user can 
access the information from another location. As shown in 
Figure 1, a new machine becomes the primary server while the 
‘heartbeat’ diagnostic program continues to check if the various 
replica servers are online and working properly. Additionally, it 
is possible to automate the failover (switching from a server 
that is down to an available one) and data recovery processes.
Fig. 1. 
Database Replication and Failover.
Scalability is the ability of databases to grow tremendously 
in size while maintaining usability or performance. There are 
very large databases that constantly accumulate data on the 
internet. In MongoDB, more servers can be allocated to a 
database to balance the workload and increase performance; 
this is called horizontal scalability. Sharding is the process in 
which a single database is divided into multiple components. 
These database components are stored on different servers. 
These servers may also be virtual machines or cloud based.
IV.
L
EARNING ACTIVITIES USING MONGODB
One of the first skills that the database engineer needs is to 
be able to quickly install the MongoDB server software on a 
given computer. A brief example in this paper is installing it on 
the Microsoft Windows operating system. After navigating to 
https://www.mongodb.org/downloads, the user needs to select 
the appropriate version from the menu, and clicks on 
‘Download MSI.’ Once the setup file is saved, the user can 
execute this, follow the instructions, and complete the steps. 
After this installation, the MongoDB server can be run from the 
Windows command prompt. However, the setup process does 
not complete all the requirements; the user needs to manually 
create a subfolder named ‘db’ in the ‘data’ folder of the hard 
drive to avoid potential error messages and run MongoDB.
Although a database engineer can interact with this basic 
installation using the command line interface, the next 
recommended and useful task is to install a front-end graphical 
user interface application to manage the databases. Figure 2 
shows the configuration of the front-end application called 
NoSQL Manager. The default port number for the MongoDB 
server is 270017.
Fig. 2. 
Connecting to MongoDB using a front-end application.
Another optional tool for database students and trainees to 
be aware of is called MongoDB Management Service (MMS). 

---