Introduction to High Availability with Amazon Elastic Load Balancers
If you are using a server-based solution to deploy and host your application, assuming that just one EC2 instance will be enough, especially a powerful one, is natural. What happens if the traffic to your application increases substantially? Upgrading your server capacity, such as increased CPU or memory, would result in an outage period, even if only for a few seconds, to scale your server with higher configuration settings. At some point, you would reach an upper limit in terms of how much you can upgrade an individual virtual machine. Furthermore, if the EC2 instance were to fail, you would need to spin up a new instance and perform necessary restoration tasks that might take longer.
Figure 7.1 – Vertical scaling has limitations
An important point to note is that vertical scaling does not provide support when the instance fails or if there is an outage in the Availability Zone...
