Developer observability is key in a world driven by the ability to fix things quickly. There are two key enablers that one needs to look at in this arena. Let’s talk about them in this article.
If you had asked me some fifteen years back about fixing issues on the fly in production, I would have said, ‘Wait, let a major release happen!’ But that is not the scenario right now in which your software is measured by performance characteristics on how quickly it can recover from a failure, because every second is important in spaces like e-commerce where downtime means loss of business in millions. One of the key metrics to measure this is the MTTR (Mean Time To Recover). As such, I am not a fan of making MTTR as the North Star for your business, as not all businesses are created equal. If you are in a space where downtime is a ‘No No’, then MTTR is important.
The other aspect about fixing issues is the ability to not to disturb code with debug statements, logging statements, etc. These statements, however effective and user-friendly they may seem, can introduce issues into the code inadvertently. You wouldn’t want to fix bugs in such statements instead of focusing on your logic, would you?
External instrumentations like plugins, that can observe code execution from outside, and being able to set observability characteristics that don’t intrude with the execution of the code are key enablers to developers and testers alike. In production environments, you won’t like to stop the system, because stopping the system means loss of business. Here is where external instrumentations help. Without stopping the system, if you could check various system parameters, variables, and context and then address the issue on-the-fly, that’s a great enabler for production.
Eran Kinsbruner and I sat down to talk about developer observability. I really liked Eran’s fast-paced way of unpacking the details about the concept, as most of the times, I would like to have my podcast session recorded within some 25 minutes, because who has the time, right? Well, it worked out well. Sit back, relax, and listen in to that podcast!
In other news, I am dabbling with multiple things in Software Security space, trying to influence quality into the way things are done in vulnerability management and threat modeling alike. More on that in the coming days and weeks, but here’s the thing, don’t forget generative-AI! It is a key thing in my radar that I’ve been closely following over various forums, and contributing to some generative-AI + security efforts.
Hey, what’s life without all this tech. fun?! But yes, I would like to share all these deep expertise with organisations that are looking for solutions for their space. So, don’t hesitate to setup a time with me to chat about your needs! Always glad to help!