You’ve probably heard, “Don’t skip leg day!”? Likewise, don’t skip monitoring and alerting for your applications.

You have monitoring and alerting in place for your cloud resources that run your money-making applications, right? RIGHHHHT??? You would probably say, “Yes, of course I do!” But then in the back of your mind, you may also think, “Well….but.”

Look, I get it. Delivering software to meet business demand can be daunting. Short timelines, shrinking teams, new technologies to be learned and implemented, sometimes having to wear multiple hats in your team. All of these pressures can cause you to skip some “nice to haves”, like monitoring and alerting, or unit testing, because you just need to deliver features because that’s what customers are asking for. Sound familiar?

While it’s true that customers do want those features, don’t overlook the fact that they also want reliability. Delivering that next competitive feature because the alternative makes your app look inferior while ignoring things like monitoring and alerting can leave your app – and your customers – in an unhappy place.

One of the worst places you can find yourself is that your customers are the ones that tell you that your application is having an issue. It’s important that you proactively monitor your application for availablity and performance. You can then step in and address things before people start flocking away from your app and go to your competitor.

Consider the observability of your application as a feature and give it the prioritization it deserves!

“I don’t know where I should start, it seems daunting!”

OK, so the first step in solving any problem is first admitting you have a problem. Admitting that you don’t have monitoring and alerting in place allows you to focus on this work.

Getting started can be done in three stages.

1. First, for all new cloud resources you spin up, determine the metrics that you want to monitor.
   • You should start with those that are availability impacting, then focus on those that are performance or user experience impacting.
   • Make an agreement with yourself and your team that you won’t spin up any new cloud resources without configuring the monitoring and alerting for them.
2. Then for your existing cloud resources, start with resources that are critical to your app such as your compute, storage, or databases.
3. Leave your less critical assets for last like cache, CDN, or other assets that your app can live without.

I like the phrase “Divide and conquer” to tackle projects like this.

Where to Go From Here

Depending on your cloud provider, I recommend you first start with one of these two links to familiarize yourself with the services available to you.

Azure – Getting Started with Azure Monitor

AWS – What is Amazon CloudWatch?

Next, I’d suggest you familiarize yourself with the term “Observability” or commonly referred to as “O11y”. It refers to taking your monitoring to the next level by exposing and analyzing insights from within your application.

When you’re ready to head in that direction, check out David Linthicum’s InfoWorld post, What Observability Means for Cloud Operations.

Go Forth and Monitor

So what are you waiting for? Dig in and start improving your ability to keep tabs on your application. Your customers will thank you (or at least not hate you).

If you found this post informative, please considering subscribing to get new posts delivered to your inbox.