Thursday, 28 June 2018

Relational Databases Advantages and Disadvantages

Relational Databases

RDBMS minimize data storage by not repeatedly the same information row by row. This is because relational database can store reference data in one table and allow other tables to reference it. Hence it is allowing a One to Many relationship through relational references Example: Students and Classes
This is not available in Flat Files and NoSQL Database Minimize Updates Example: Let say the student address changes but classes stay the same. Because every row has a student address the whole data must be updated accordingly.
RDBMS ensure data integrity. Not available in Flat Files and NoSQL
RDBMS ODBMS ensure data is kept central live. Not available in FlatFiles multiple copies.
RDBMS ODBMS ensure data is transactional and accurately tallied. Not available in FlatFiles and NoSQL.
Example: School Fees and Credit Card Payment Tallied correctly even if one fails. Flat Files

Relational Databases

RDBMS does not maintain integrity rules over specific types of data away from the data storage layer. ODBMS does. Example: Object Oriented Relational Database can specify English / Mathematics / Humanities as Classes (Subjects) away from another Class like ART. RDBM does not distinguish this through its structure. They are both though accurate and centrally located data stored in centrally located server.
RDBMS Integrity makes it difficult to import 3rd part non-conforming data. Thus data integrity is enforced rigorously in RDBMS. Same for ODBMS. FlatFiles and NoSql does not enforce this rule.
RDBMS Performance Slow for storing data like html and is slow than transforming to table format and does grow the database making it slow down. As a result Indexes need to be optimised for inserts and updates and viewing. NoSQL does not have a performance hit and is very fast. So are Flat Files.
RDBMS Table(s) Requires programs to be written to display data in a user friendly manner to display business information. FlatFile is raw format of the data and software already available. NoSQL and ODBMS has the same limitation.

As a result, the main reason for RDBMS is for real time live data inserts, updates and viewing. Also, when interfacing to other systems RDBMS accomplishes this to be done with full data integrity in real time using Transactions

No comments:

Post a Comment