The state of error monitoring in 2022
In the modern technical world, we are less plagued by the blue screen of death
…and more so by slow performance, UI glitches, and data availability.
When users are using your product, whether it’s an API, web app, or mobile app, how can you tell what users are experiencing without tools to understand your infrastructure performance?
With complex systems, bugs can be difficult to spot and complicated to solve.
The way applications are designed today, there is less of: track all hard crashes and send a dump to the developer to pick through and more of: your servers are slow.
With this complexity and ambiguity, how do we bridge the gap between an engineering organization and its users? When a bug is found, how do you triage and prioritize it?
Error monitoring today
Products of today enable error monitoring that can help teams make informed decisions about bugs–how many users it affects, how often it occurs, and when they started showing up.
These questions can be answered with tools like Datadog or Raygun and can unify your system’s logs to give a time-stamped view into your system.
You also want to leverage your company’s internal messaging system (e.g., Slack) to alert relevant parties about bugs if you have an event system. Alerting your software engineers when messages entered the DLQ helps them know to go in and inspect.
How to choose an error monitoring product
That’s a tricky question, as always it comes down to it depends.
But make sure you approach the discovery phase with clear outlines. Some things that are important when trying to decide what error monitoring product to use are:
– How out-of-the-box is it? Most often, you want something that can easily be added to different services and apps without too much tinkering
– Does it provide API integrations for things like Slack notifications?
– Can you easily find the full stacktrace if there was a critical error?
– Can you find performance issues with it?
– Does it support the technologies I will be using? ( Web/mobile/backend?)
– What does the pricing structure look like?
Understanding what is essential to your organization and what kind of issues you want to alert on will help you decide which one to use.
Generally, most have a free tier, so when you’re in a proof of concept/discovery phase, you can use a couple of different ones to see how they compare.
The importance of error monitoring
Error monitoring can help you gain better insights into your deployment process. Being able to identify a lousy build quicker and respond to it, even before it gets to users, utilizing strategies such as canary deployment with real-time performance and error tracking can help you make deployments more confidently.
There is no such thing as bug-free code. But having a plan in place to address them is key to success.
Personalize your experience