The Why of Distributed Systems

Why do we need distributed systems?

We have already kind of answered this question in our previous lesson. But let’s look at the “why” in detail.

There are quite a few reasons why we need distributed systems:

• As you’ve seen in the last lesson, a non-distributed system runs on a single node, and that node always has a chance of crashing. Actually, nodes crash all the time in large systems, for example, at Google.
• The above means that it’s best to keep multiple copies of your server and database running on different nodes, and these nodes should be consistent in terms of functionality.
• With time, your system may grow. Say your My Cool App started with no user, and after six months it was so successful that you now have 1 million monthly active users. When you had no user, running the app on a single server made sense, but not after six months. Your server program needs to be running

...