Home

DemoKafkaAkkaSpark

Demo of RSS management and Reporting with Kafka Akka and Spark

Needs

• Spark installed on directory spark-2.3.1-bin-hadoop2.7
• SBT installed
• Python with modules : yaml json lxml bs4 kafka
• Kafka started see https://hub.docker.com/r/wurstmeister/kafka/

See the Presentation

https://github.com/buidiaconseil/DemoKafkaAkkaSpark/blob/master/KafkaAkkaDemoPres.pdf

Generate content.rss

Get the content from the web and push the resulted xml into the file content.rss

python pullRSS.py

Push the content of content.rss

We put in json the file content.json

python pushRSS.py

Demo en Akka Stream

In the directory akkastream

5 demos:

• demo0 : consume the flow and get it on console
• demo1 : consume and get just the words on console
• demo2 : groupBy parallel of demo0
• demo3 : groupby + merge
• demo4 : A complex flow to calulate an iterative TFIDF

to test launch on the directory akkastream:

sdb "run demo0"
sdb "run demo1"
sdb "run demo2"
sdb "run demo3"
sdb "run demo4"

Demo on Spark

Spark installed on directory spark-2.3.1-bin-hadoop2.7 From the root directory , launch

sparkStarterSup.sh -i kafkaSpark.scala

This first demo launch a consumer and get every 5 seconds the json description

sparkStarterSup.sh -i kafkaSparkTFIDF.scala

This first demo launch a consumer,get every 5 seconds the json description, push it on a memory table and we launch manualy the calculation of the TFIDF