Before you say, "Shahid, we have learned this in previous chapters, why do it again now?", I want to hold you up there. I know we did cover this topic in our previous chapters, but I would like to revisit it, because it's really important for your application deployment plan.

If you are a DevOps, you can thank me later (they know what I mean).

Cosmos DB provides a competitive feature, which none of the other managed database providers, such as Amazon, Rackspace, and so on, offer, at least at the time of writing. Hence, it becomes a very important selling point of Cosmos DB.

Let's revisit it again and see how it can help your deployment plan. So, you can replicate your data across more than 30 regions on the planet with just a few clicks of a button (not really one click, but very simple). That's what turnkey global distribution means.

So, a possible use case in the deployment plan could be as follows. Suppose you have an SaaS-based application and you have a huge customer base in Germany, but your application is deployed at a New York data center. Now, based on the data transfer rate, you can spin up one read region in Germany (if they provide one there) or near to Germany to facilitate those customers in a much better way.

Mainly, the replication is meant for the failover conditions; if some node fails in a particular region, then another node with the same data takes over and customers never know that anything has changed.

For example, refer to the following screenshot from Chapter 2, Cosmos DB Global Turnkey Distribution System:

Here, my write region is based in Southeast Asia and I can pick and choose any read region from across the globe. There are tons of data centers in the states, very few in Russia, and none in Africa (which isn't fair; Africa is blooming in IT).

All you need to do is pick a region and see the magic happen; you get your replication done.

With this mention of replication, let's dive into the next important topic of deployment, that is, failover handling.