Hey Product Hunt! 👋 We built Umbrelog after one too many 3am incidents where the logs were right there — but disconnected from everything that mattered. No link between the error and the DB call that caused it, the deploy that shipped an hour earlier, or the queue message that triggered it all. At the same time, every log tool we tried had the same second problem: pricing that made no sense until the bill arrived. One extra host, one debugging spree, and suddenly costs doubled. So we built Umbrelog around two things: automatic correlation (logs + runtime context + deploys, reconstructed automatically) and pricing you can actually predict in advance. Still early days — would love your feedback, especially from anyone who's been burned by unpredictable observability bills. 🙏

Forum Threads

p/umbrelog

•

2mo ago

What do you use when logs aren't enough?

When something breaks in production, the logs usually tell you what failed but figuring out why can mean jumping between logs, deployments, databases, queues, and metrics.

Curious how other engineering teams handle this today.

What s your workflow when you need to understand what actually happened during a production issue?

View all

Umbrelog

Turn production logs into answers

Turn production logs into answers

Forum Threads

What do you use when logs aren't enough?

Forum Threads

What do you use when logs aren't enough?