In OpenShift application, auto SCALING is also known as pod auto scaling. There are two types of application scaling as follows.

• Up (vertical scaling)
• Out (horizontal scaling)

 Imagine you have a successful start-up and your user base is growing. As the user base expands, application load and demand increases. To keep up with this demand you either need to add servers (horizontal scaling) or get bigger ones (vertical scaling).

Vertical Scaling: To accommodate higher load using vertical scaling, your application REMAINS in ONE place and you give it more resources. For example, you might add a larger machine with more CPUs, faster CPUs, memory, or disk space. Cost CONTINUES to rise as you add more HARDWARE resources.

Horizontal Scaling: To accommodate higher load using horizontal scaling, multiple instances of an application are created and the application load is balanced across independent nodes.

This has several advantages because you can:

Use ordinary hardware and thus keep the price of your hardware use reasonable
Deploy hundreds or thousands of nodes and load balance the application between them

In OpenShift application, auto scaling is also known as pod auto scaling. There are two types of application scaling as follows.

 Imagine you have a successful start-up and your user base is growing. As the user base expands, application load and demand increases. To keep up with this demand you either need to add servers (horizontal scaling) or get bigger ones (vertical scaling).

Vertical Scaling: To accommodate higher load using vertical scaling, your application remains in one place and you give it more resources. For example, you might add a larger machine with more CPUs, faster CPUs, memory, or disk space. Cost continues to rise as you add more hardware resources.

Horizontal Scaling: To accommodate higher load using horizontal scaling, multiple instances of an application are created and the application load is balanced across independent nodes.

This has several advantages because you can:

Use ordinary hardware and thus keep the price of your hardware use reasonable
Deploy hundreds or thousands of nodes and load balance the application between them