This demo processes Meetup.com RSVP stream and stores results in Cassandra. The main purpose of this application is to demonstrate Spark Streaming capabilities in Spark Cassandra Connector and DSE.
It produces two types of results:
- Total number of attendees per country (since application start). Refreshed every 5 seconds. Shows simple stream processing functionality.
- Trending meetup topics within the last 5 minutes. Refreshed every 10 seconds. Shows windowed transformations functionality.
Web dashboard is included for visualisation of these results:
To start stream processing locally:
git clone [email protected]:rstml/datastax-spark-streaming-demo.git
cd datastax-spark-streaming-demo
sbt assembly run
To deploy Spark application on DSE cluster:
dse spark-submit --class com.datastax.examples.meetup.StreamingDemo ./target/scala-2.10/streaming-demo.jar -Dspark.cassandra.connection.host=127.0.0.1
Input options:
- -Dspark.master - Autodetect in DSE. Specify for non-DSE deployments.
- -Dspark.cassandra.connection.host - Default is 127.0.0.1, replace with rpc_address of one of the nodes.
- -Dspark.cores.max - Default configured is 2
- see resources/applicaiton.conf for more
To start web applicaiton:
cd web
./sbt
> container:start
Point your browser to localhost:8080 and watch map and topics update in real time.
Other endpoints:
- /countries - attendees by country since start
- /trending - trending topics within last 5 minutes
To deploy app to a servlet container, create war package using command below:
cd web
./sbt
> package
This software is available under the Apache License, Version 2.0.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent