Distributing Load with Azure Traffic Manager

Sometimes we want to distribute our load depending on the performance of our backends, or maybe route users to different servers while some are under maintenance. This is not an easy task if we don't have a service that will do this seamlessly and quickly. Thanks to Azure Traffic Manager we are able to improve the availability of our critical applications, distribute traffic when performing large, complex deployments, or perform maintenance without downtime.

The following topics will be covered in this chapter:

• Using Azure Traffic Manager
• Different routing methods
• Endpoint monitoring

Technical requirements

To perform exercises from this chapter, you will need the following:

• Access to an Azure subscription

Azure Traffic Manager fundamentals

Imagine the following situation—you have an application that has to be served globally. To guarantee the best performance for all your customers worldwide, you provision different instances of your service in different regions (one for North America, one for Europe, and one for Africa). There is one problem, however. You have to explicitly tell your customer to access a specific instance of the application—the one that is closest to its location.

While this is, of course, possible (just give it the right URL), the solution is not ideal. For example, what if your client goes for a holiday and spends the following two weeks in Europe instead of in Africa? To overcome such problems, in Azure you can leverage a service named Azure Traffic Manager, which takes care of the proper routing of incoming requests and allows you to implement...

Monitoring

Besides routing traffic to a different endpoint, Azure Traffic Manager offers some additional functionalities when it comes to monitoring. Besides the traditional Metrics blade, there is an extra feature available called Traffic view, which enables you to monitor. What is more, you can use many different built-in mechanisms (like nslookup in Windows OS), to check the current configuration of the service.

Nslookup

To use nslookup, you have to run the command line in Windows using your administrator account. Once it is loaded, enter the following command:

nslookup <Traffic-Manager-DNS-name>

After a moment, it should return a result showing the command resolution:

DNS request timed out.
    timeout was 2 seconds...

Summary

This was the last chapter in this book and explained the basics of one Azure services, Azure Traffic Manager. You have learned the fundamental concepts of traffic distribution and different routing methods that cover many real use cases, which you may well face in your daily work. Now you should understand how this particular Azure service works and what can be achieved by the proper usage of its features, such as configuration, real user measurements, and monitoring. In the next (and, unfortunately, the last) chapter, I will show you some useful tips and tricks for working with the Azure Portal and different cloud components to improve your skills even further.

Questions

1. What are the supported routing methods in Azure Traffic Manager?
2. How can you use the Real user measurements feature?
3. Can you link different Azure Traffic Manager profiles?
4. Is it possible to use an external endpoint?
5. Does a client connect directly to an endpoint returned by Azure Traffic Manager?
6. What is the main difference between a gateway and Azure Traffic Manager?
7. Can Azure Traffic Manager be used to achieve high availability? If so, how?

Further reading

• Disaster recovery with Azure DNS and Traffic Manager: https://docs.microsoft.com/en-us/azure/networking/disaster-recovery-dns-traffic-manager
• How it works: https://docs.microsoft.com/en-us/azure/traffic-manager/traffic-manager-how-it-works