Data Modeling And SQL Queries For A Library Database

Overview of the Schema

2 a) Relational Database Schema

TblPublisher (Fldname: text, Fldaddress: text, Fldphone: text, FldURL: text)

TblAuthor (Fldname: text, Fldaddress: text, FldURL: text)

TblBook (FldISBN: number, Fldtitle: text, Fldprice: currency, Fldyear: number, FldauthorName*: text, FldpublisherName*: text)

TblCustomer (Fldemail: text, Fldname: text, Fldaddress: text, Fldphone: text)

TblShoppingBasket (FldbasketID: number, FldcustomerEmail*: text, FldISBN*: number, FldNoOfBooks: number)

TblWarehouse (Fldcode: text, Fldaddress:text, Fldphone:text)

TblStocks (FldwarehouseCode*: text, FldISBN*:number, FldNoOfBooks: number)

(Janalta Interactive Inc 2015)

2 b) Composite Attributes

The composite attribute can be separated into two or more meaningful attributes.

The BookStore database has following composite attributes-

Address – an address attribute may be separated into StrretAddress, City, State, Postcode attributes. So, an Address is the composite attribute.
Name – a Name attribute may be separated into FirstName, LastName, MiddleName attributes. So, a Name is the composite attribute.

3 a) Hadoop

Introduction

Hadoop is an open source system which is used to store up and process a Big Data. It works in distributed environment with number of systems as cluster systems. “MapReduce” algorithm is developed by the Google to store the Big Data and the Hadoop is based upon the “MapReduce”.

It is very difficult to process the Big Data but Hadoop has resolved the problem to process the big data. It splits the big task into smaller tasks and distribute to number of systems to process. This is called Map and after processing it collects the results from all the systems and generates final result to show to the user.

Capabilities

Hadoop is very capable in the following fields-

It is compatible on all platforms.
It is very flexible. A server can be added or removed dynamically, but the Hadoop doesn’t stop. It works in the same manner.
Hadoop library is very fault tolerance. It detects the fault. It does not rely on any hardware to detect the fault.
It is very inexpensive.

Limitations

Hadoop has some limitations like below-

It is not appropriate for small data. It works well on Big Data.
The batch processing is only supported in Hadoop.
The processing speed is very slow in Hadoop because of Big Data and MapReduce algorithm.
It is not applicable on real time applications as it does not support the small data.
It is not proficient in iterative processes.

Conclusion

Hadoop is the best technique to process the Big Data. It does not suit on small data but works very well in Big Data as it is only made to resolve the issues that are being occurred in the Big Data processing.

(TechTarget 2018)

3 b) MapReduce

Introduction

“MapReduce” is used to make applications to process Big Data. It is the software framework. Hadoop is also based upon MapReduce. It makes the applications which works in distributed environment in parallel.

It divides the large task into smaller tasks and distribute to the number of systems to process Big Data. It assigns the small tasks to the parallel systems that are connected into distributed environment. After processing from each system; it sorts all the results and combine them to produce final result. Reduce works in the following steps- shuffle, Sort and Reduce. It works in the pair of <key, value>.

Capabilities

MapReduce is very capable in the following fields-

It is very scalable. It is independent of any environment.
It is not so much expensive.
It may work on structured or unstructured data both.
It is very secure and works with the HBase security.
The data in copied for backup when the data comes to process at the main node. The MapReduce works on this concept. It is very helpful in backup techniques.

Limitations

MapReduce has some limitations like below-