tracylu2015 / news-streams

News-Stream Android Mobile App

news-streams project

news_stream_django

1. Required environment variables
   • MONGODB_URI
     • A Mongodb Uri with username, password and mongodb dns
   • MEMCACHED_SERVICE
     • A Memcached service DNS name which include all Memcached IP addresses
     • Client can resolve this DNS name to get all IPs in this Memcached cluster
   • ELASTIC_HOST
     • An Elasticsearch local host to get search results in Elasticsearch database
2. How to run after setting environment variables
   • direct to news_stream_django folder
     • run command ''' python manage.py '''
3. Endpoints
   • /api
     • to get all twitters information according to SocialPost model including required hashtags
     • Return format: Json
   • /api/trending
     • to get all recently mentioned top counted hashtags aggregate from mongodb
     • Return format: Json
   • /api/tags/{tag}
     • to get tweets which text including specified tag
     • param need to pass: string of tag pass from frontend url
     • Return format: Json
   • /health
     • to check whether app is health
   • /metrics
     • to get kafka received tweets and processed tweets

news_stream_frontend

• React command to run on Android ''' npm run android

'''

news_stream_streams

1. Required environment variables
   • CONSUMER_KEY / CONSUMER_SECRET / ACCESS_TOKEN / ACCESS_TOKEN_SECRET /BEARER_TOKEN
     • twitterAPI credentials
   • KAFKA_BOOTSTRAP_SERVERS
     • a list of host/post pairs separated by "," to establish initial connection to the Kafka cluster
   • KAFKA_GROUP_CONSUMER
     • a Kafka consumer name which cooperate multiple consumers to consume data from same topics
   • MONGODB_URI
     • A Mongodb Uri with username, password and mongodb dns
   • ELASTIC_HOST
     • An Elasticsearch local host to get search results in Elasticsearch database
2. How to run after setting environment variables
   • direct to news_stream_stream folder
     • run command to get Twitter stream and produce raw data to Kafka

       python twitter_stream.py
       

     • run command to start consume and process data and insert into database

       python kafka_main.py
       

3. streams (from twitter and reddit)
   • filtered twitter posts according to twitterAPI rule
   • add kafka producer to the filtered tweets and prometheus metrics
   • add kafka consumer and prometheus metrics before insert data into mongodb database
4. models
   • SocialPost model to parse required fields from twitter posts with related tags
5. mongo_query
   • mongo aggregation to have top counted hashtags and converted to python based
   • merge same categories of hashtags into one list and sum counts
   • add to memcache to reduce database queries times

Dockerfiles

Dockerfile for both streams and django service which can be built in CI to docker image

Deploy

Contains Kubernetes configurations for all components which includes: - backend-api - memcached - news-streams