Observability overview

Unified observability for commerce extensibility.

Over the last few years, extensibility in Adobe Commerce has evolved into a clearly composable model. Today, every app is independent. Each extensibility model has their own execution model, and integration evolves on their own pace. From an architectural perspective, this means that a single commerce experience can span multiple JL layers. Webhooks reacting to a platform event, event subscriptions triggering asynchronous processes, UI extensions executing client-side logic, and App Builder apps orchestrating business workflows. Each of these components is isolated by design, which is exactly what makes the model scalable and flexible. But it also means that there is no single execution context by default. If you look at architecture, it becomes clear that tracking what happens across the system cannot rely on individual logs or isolated metrics. To understand the system behavior, you need a centralized observability layer capable of correlating signals across all these independent components. That correlation layer is what allows you to reconstitute a complete execution flow, even when that flow spans multiple apps, services, and extensibility components. At this point, observability stops being an implementation detail and becomes an architectural requirement.

Now let’s bring that to a real-life example. Imagine a shopper at the checkout. When they click on the place order, the system triggers multiple webhooks. One to validate the payment, another for tax, and another for inventory availability. All these costs happen in real-time, and any delay or failure directly impacts the shopping experience, and ultimately, conversion rates. Without observability, identifying what went wrong is a guesswork. Was the delay in the tax service? Was the inventory validation too slow? Or did a webhook fail regularly? With observability signals in place, logs, traces, and metrics all correlated, the merchant can see exactly which service is causing the bottleneck, and act immediately, rerouting the traffic, or back up a service, disable a problematic extension, or just adjust time-outs to preserve checkout responsiveness. This level of visibility transforms troubleshooting from a reactive firefighting into a proactive optimization, ensuring every checkout is fast, reliable, and conversion-friendly.