Riemann monitors distributed systems.

Riemann streams are just functions which accept an event. Events are just structs with some common fields like :host and :service You can use dozens of built-in streams for filtering, altering, and combining events, or write your own.

Since Riemann's configuration is a Clojure program, its syntax is concise, regular, and extendable. Configuration-as-code minimizes boilerplate and gives you the flexibility to adapt to complex situations.

I wrote Riemann for operations staff trying to keep a large, dynamic infrastructure running with unreliable but fault-tolerant components. For engineers who need to understand the source of errors and performance bottlenecks in production. For everyone fed up with traditional approaches; who want something fast, expressive, and powerful.
See problems faster

Traditional monitoring systems run polling loops every five minutes, or roll up metrics on a minutely basis. In a Riemann infrastructure, clients (including stand-alone pollers) *push* their events to Riemann, which makes them visible within milliseconds. Low latencies let you see outages faster--and know the instant you've fixed the problem.

Throughput depends on what your streams *do* with events, but a stock Riemann config on commodity x86 hardware can handle *millions* of events per second at sub-ms latencies, with 99ths around 5ms. Riemann is fully parallel and leverages Clojure and JVM concurrency primitives throughout.

http://riemann.io/