Caudal is a data analysis platform that receives application events. Using a powerful DSL, brings you a sophisticated way to enrich and correlate all incoming events, in order to get rich information for reliability monitoring.
They are a mechanism to put data in Caudal. Anything can be a datasource: application logs, a Telegram chat, or a Twitter feed.
Functions than receive as input some data and returns structured events.
All data in Caudal are received in form of events. An event are any data struct and is passed as a Clojure Map.
A streamer is function applied for each incoming event and can be composed and combinated to enrich the data stream.
Caudal uses an application State in order to store data needed for metrics, statistics, configurations, etcetera
The following file shows a basic configuration that captures data from TCP 9900 port, encapsulates this data in a map, then passes it to two streamers, the first one counts event and the second prints the enriched event in log.
;; Requires
(ns caudal.config.basic
(:require
[caudal.io.rest-server :refer :all]
[caudal.streams.common :refer :all]
[caudal.streams.stateful :refer :all]
[caudal.streams.stateless :refer :all]))
;; Parser
(defn basic-parser[incoming-data]
{:message incoming-data})
;; Listener
(deflistener tcp [{:type 'caudal.io.tcp-server
:parameters {:parser basic-parser
:port 9900
:idle-period 60}}])
(defsink example 1 ;; backpressure
;; streamer
(counter [:state-counter :event-counter]
; streamer
(->INFO [:all])))
;; Wires our listener with the streamers
(wire [tcp] [example])Running this configuration:
