Scale Out vs Scale Up - to be renamed accordingly

Let us discuss the terminologies scale out and the scale up in this article - in general and also on the context of Big Data.

Though these terminologies are mostly inclined with the Big Data, they are still applicable to the outside as well for the Scaling.

Scaling

Generally, the term Scaling indicates the ability to accommodate the increased demands of serving the clients for the Software/Application or the Data. The scaling can be accomplished at various aspects of the System/Server - be it in terms of increased memory (RAM) , hard disk, additional machines / systems in the same network topology etc., depending on the actual type of demand.

Types of Scaling Functions

We have heard the terms of ‘Horizontal’ and ‘Vertical’ scaling and they indicate the nature of scaling we perform.

• Horizontal Scaling - indicates we add the capacity in the horizontal manner - to accommodate them in more number of parallel servers - which indeed forms a chained connectivity from one node to another - expanding in the horizontal direction (left to right).
• Vertical Scaling - indicates the ability to add the capacity in the same server - by adding on top of one another - thus forming a vertical stack.

Scale Out Vs Scale Up

The terminologies Scale Out and Scale Up - by now you would have imagined/predicted that they are the interchangeable terms of the Horizontal and Vertical scaling respectivley.

Big Data Context

These terminologies fit well in the Big Data Context - where in Hadoop, the nodes (HDFS - Hadoop Distributed File Systems) - facilities more nodes to be added for processing more chunks of data in the series (Scale Out), thanks to the commodity hardware available nowadays at a very afforadable cost. Where needed, they can be added on the same physial server, on top one of another - where it is called as Scale Up.

Cheers,
RM…
Raghavan alias Saravanan Muthu
05 Dec 2021 | Sun | 21:52:08 PM IST