# Read via REST API from Kafka topic # Prerequisite: set up kafka with topic accesslogapache library(httr) library(jsonlite) library(dplyr) kafka_rest_proxy <- "http://localhost:8082" consumer <- 'my_json_consumer' consumer_instance <- 'my_consumer_instance' topic <- 'accesslogapache' # Create consumer # See https://docs.confluent.io/current/kafka-rest/docs/intro.html#produce-and-consume-avro-messages # Create consumer response <- POST(url=paste(kafka_rest_proxy, "consumers", consumer, sep="/"), content_type("application/vnd.kafka.v2+json"), accept("application/vnd.kafka.v2+json"), body=paste0('{"name": "', consumer_instance, '", "format": "json", "auto.offset.reset": "earliest"}') ) fromJSON(content(response, "text")) # Subscribe it to topic response <- POST(url=paste(kafka_rest_proxy, "consumers", consumer, "instances", consumer_instance, "subscription", sep="/"), content_type("application/vnd.kafka.v2+json"), body = paste0('{"topics":["', topic, '"]}') ) response # Obtain all (or latest) messages on the topic messagesJSON <- GET(url = paste(kafka_rest_proxy,"consumers",consumer,"instances",consumer_instance,"records", sep = "/"), accept("application/vnd.kafka.json.v2+json"), encode="json") messages <- fromJSON(content(messagesJSON,"text")) createPlot(messages$value) Sys.sleep(120) # Obtain latest messages messagesJSON <- GET(url = paste(kafka_rest_proxy,"consumers",consumer,"instances",consumer_instance,"records", sep = "/"), accept("application/vnd.kafka.json.v2+json"), encode="json") messages2 <- fromJSON(content(messagesJSON,"text")) apachelog <- rbind(messages$value,messages2$value) createPlot(apachelog) # Remove the consumer DELETE(url = paste(kafka_rest_proxy,"consumers",consumer,"instances",consumer_instance, sep = "/"), content_type("application/vnd.kafka.v2+json"))