Relational V S Key Value Stores Information Technology Essay

RDBMS is a database direction system which describes the information as a relational theoretical account. It defines the information and the dealingss in signifier of tabular arraies. Tables consist of rows and columns. Relational theoretical accounts have a information theoretical account in topographic point along with assorted defined restraints. SQL question is used to recover informations from the database. SQL includes a batch of complex questions which involves unity restraints and articulations.

Key-value distributed shops allows storage as a simple hash tabular array. It shops keys and the keys map to a value. The hunt is conducted on the keys and it returns the value. The value is either stored as binary object or semi-structured like JSON.

Need essay sample on Relational V S Key Value Stores... ?We will write a custom essay sample specifically for you for only $12.90/page

order now

“ SQL databases are like automatic transmittal and NoSQL databases are like manual transmittal. Once you switch to NoSQL, you become responsible for a batch of work that the system takes attention of automatically in a relational database system. Similar to what happens when you pick manual over automatic transmittal. Second, NoSQL allows you to eke more public presentation out of the system by extinguishing a batch of unity cheques done by relational databases from the database grade. Again, this is similar to how you can acquire more public presentation out of your auto by driving a manual transmittal versus an automatic transmittal vehicle. “ [ 1 ]

Key-Value shop is portion of the NoSQL community ( Not Merely SQL ) .




Recovering and hive awaying informations may necessitate multiple tabular arraies

They store informations about a peculiar point ( which is used as the key ) along with that point

Defines the theoretical account and application is developed to map to the theoretical account

Application-driven theoretical account in the sense that the application needs to make up one’s mind the theoretical account for hive awaying and recovering the information

Data unity and restraint cheques are enforced by the defined theoretical account

Data unity and restraint cheques have to be enforced at the application-level

Data is accessed by utilizing a query linguistic communication like SQL

Data is accessed by utilizing the API of the key-value shop

As the information theoretical account and concern logic are independently developed, the same informations theoretical account can be re-used for a different application

The information in the shop is application-specific and the opportunities of reuse are less as compared to relational theoretical account



We will hive away the information about books in a tabular array. See the books table as below:





A Whole New Mind

Daniel Pink


The Black Swan

Nassim Taleb


Power of Now

Eckhart Tolle

We will compose the question in SQL as follows:

Query 1: To acquire all books from a peculiar writer

choice * from books where author_name = “ Daniel Pink ” ;

Query 2: To acquire all writers

choice author_name from books ;


See that we have to run the same questions for a key-value database. We will hold to hive away the information in 2 key-value braces as below:

writers = & gt ; { “ Daniel Pink ” , ” Nassim Taleb ” , “ Eckhart Tolle ” }

writer: Daniel Pink = & gt ; { “ A Whole new head ” , “ Drive ” }

We will compose the question utilizing the key-value shop ‘s API as follows:

Query 1: To acquire all books from a peculiar writer

get ( writer: Daniel Pink ) ;

Query 2: To acquire all writers

get ( writers ) ;

The logic for seeking is pretty complex in key-value but they are manner faster than database.


There are some relational database operations that are a public presentation hit:

Changing a tabular array consists of:

Updating the scheme

Inserting or canceling values for the ( new/old ) rows and columns

It may take hours to finish this operation for 1000000s of rows. While executing such operations, the database tends to lock the tabular arraies and this consequences in a public presentation hit.


Articulations are comparatively slow. They are one of the chief grounds of utilizing relational theoretical accounts. Excluding articulations would be every bit good as non utilizing relational databases.

Consequences of MySQL, Redis and Tokyo Tyrant public presentation comparing [ 2 ]



Relational databases tend to scale good when the full database is located on a individual waiter. The job starts when the individual waiter can no longer manage the petitions and the information has to be partitioned across waiters.

One of the manner to make that would be to utilize reproduction. The job with that it that it takes clip for the reproduction to ab initio sync with the maestro and be on par with the database. This might be a immense job if the system is a societal networking site and is used by tonss of users.

Another facet is to shard the database. One of the techniques of database sharding involves horizontal breakdown. It fundamentally divides the rows of a tabular array to do a new divider. Each divider is called a sherd.

Idaho | name

| Daniel Pink

| Nassim Taleb

Idaho | name

| Daniel Pink

| Nassim Taleb

| Eckhart Tolle

| David Hansson

Idaho | name

| Eckhart Tolle

| David Hansson

Problems with Database Sharding

You will hold to shard the database maintaining in head the entree forms of the end-users

Besides, sharding agencies that you can non run questions that involve fall in operations across sherds as this would be inefficient

Using referential unity across databases is besides a job in sharding the database

It ‘s a complex system to implement in pattern

So, basically sharding the database means that you lose all the advantages of the relational theoretical account.

Key-Value databases can scale truly good as it ‘s non a individual point of failure and can digest multiple failures. Addition of new nodes can besides be done easy and thereby helps to do the application scalable.

Application particulars

Applications need to make up one’s mind if they have to back up more characteristics or grading.


“ The CAP theorem, besides known as Brewer ‘s theorem, states that it is impossible for a distributed computing machine system to at the same time supply all three of the undermentioned warrants ” [ 3 ] [ 4 ]

Consistency provinces that every client has the same version or position of the informations

Availability states that the service will ever be able and read and write can be performed on the waiter

Partition tolerance means that the system can work even if some of the nodes go down

Visual usher to NoSQL [ 5 ]

Relational Database

Consistency is the nucleus of relational theoretical account and it is the ground for its monolithic use in applications.

If you design your database for consistence and handiness so if the node fails there is no failover

If you design by maintaining consistence and partition-tolerance in head so it means that the informations would non be available till the failover node gets to a consistent province in instance of any failure

Key-Value shops provide Availability and Partition tolerance at the cost of Consistency

You can follow a theoretical account depending on the demands of the application. If you are utilizing a banking application, consistence is the most of import facet and you have to utilize a relational database.

If you are constructing a societal networking web sites used by 1000000s of users, it ‘s of import that the information is available to the users instantly. In that instance you can choose for a Key-Value database.

Advantages of relational database

Relational databases support minutess and all ACID belongingss. These are indispensable for concern and critical applications to ever hold consistent informations Examples: paysheet or banking applications

RDMS tools are mature and have been tested over a period of clip to work out assorted sorts of jobs

As SQL is a standard linguistic communication used by all the relational database sellers with comparatively minor alterations in execution, you have the flexibleness to travel to a different merchandise as most of them use SQL for questioning the database.

It considers hive awaying every bit good every bit seeking as the same job

Disadvantages of relational database

The developer has the operating expense of doing certain that the informations theoretical account maps to the concern logic

Relational database is non distributed by nature and shops all the informations in one waiter and this acts as a individual point of failure

Besides in instance of monolithic deployments, it requires specialized hardware such as SAN ‘s to scale it

Relational databases are non scalable

Advantages of Key-Value shops

It does non hold any scheme and so you are non bound by the theoretical account

Simple questions like get, put and delete in most instances allows the system ‘s public presentation to be a batch more predictable

It can digest failures depending upon the constellation

It can run on less-expensive hardware

High public presentation as compared to relational databases

Easy to scale by adding an excess nodes

There is no demand of a DBA to pull off the database

Disadvantages of Key-Value shops

Generating dynamic studies from informations is hard

There is no easy manner to export informations from the shop

Migrating from one merchandise to another merchandise is hard in as the API used to entree informations is store specific



Get your custom essay sample

Let us write you a custom essay sample

from Essaylead

Hey! So you need an essay done? We have something that you might like - do you want to check it out?

Check it out