In cloud computing, horizontal and vertical scaling are used to deal with increased workloads and improve system efficiency.
Horizontal scaling (called scaling out) includes adding new machines or instances to spread a load. By utilizing this method, the capacity of equipment is increasing through distributing data processing on several servers, and is the preferred option in situations where fault tolerance and high availability is of essence since it allows for both redundancy and load balancing.For example, when a web application is overwhelmed by increased traffic, what it can do to cope with the pressure is increase the number of web servers to take care of the surge. This will make sure that there is smooth flow for its users without overloading the main server.
Another way is vertical scaling which involves increasing the capacity of certain computer parameters in a single computer (CPU, RAM, disk space). It is easy to understand and does not involve managing different servers since all you need is more capacity on one machine.For example, when data demands increase on a database server, the need for more powerful instance increases in order to ensure efficient processing of complex queries, and provide increased memory.
Both scaling methods have distinct requirements; on the one hand, horizontal scaling results in scalability along with resilience, and vertical scaling leads to its simplicity as well as improvement of power for resource hungry applications.