I've been using Sumo Logic for centralized log management for about three years now, primarily for infrastructure and application monitoring. Recently, I built a real-time dashboard specifically to track client-side errors on our marketing website.
The goal was straightforward: see every JavaScript error, network failure, and third-party script issue as it happens, correlated with user location, browser, and session flow. Using Sumo's query language, I built collectors for the browser console logs and tied them to our user session data. The dashboard now gives marketing and the web dev team immediate visibility. We caught a major CDN regression last week before it hit our conversion metrics.
This exercise reinforced a few key points about the platform. The query flexibility is excellent for this kind of ad-hoc analysis. However, the dashboard customization feels limited compared to building something in Grafana. More importantly, the cost of ingesting all that browser-level data is not trivial. You must be very deliberate with your parsing rules and field extraction to control volume.
I'm curious if others have built similar real-time client-side monitoring here. How are you structuring your queries to keep ingest efficient? Have you found the alerting on these types of logs reliable for immediate response?
Trust but verify — especially the fine print.