Twitter-Akka

Description:

Akka Actor based client server model to support millions of users and millions of Tweets
A full feature twitter simulator
An in memory server to process client requests
Web frontend with REST APIs implemented in Spray Can
Docker compatible

Following are the individual feature lists

Client(Simulator) :

Support customizable user Profiles.
The user profile has parameters like :
- Number of tweets per day,
- total user count contribution ( number of users of a given profile/total user count)
- portion of this user profile that retweet
- average follower count
- average following count
- average user timeline refresh rate
- average home timeline refresh rate
- average mention timeline refresh rate
Create millions of users with varying profiles
create users and followers
pull timelines for each user at a given interval. We will first implement a pull mechanism to fetch timeline. Each user client actor will ask server for timelines based on the last fetch time.
Record client statistics

Server :

Register Users
Authenticate and Login user
Recieve tweets from user
Process incoming tweets :
- save tweets in the usertimeline of the user,
- save tweets in the home timeline of the followers
- save tweets in the mention timeline of the people mentioned in the tweet.
- save tweets in the hashtag timeline
server respective timelines to user based on the last refresh time.
- The client will request server for various timelines and server will send only the new tweets which are not yet delivered to client.
Record server statistics, performance counters.

Performance Server :

this componenet/Actor/Client will gather performance statistics from the main server
display the gathered performance counter on screen at fixed intervals

Also, this project should be tested with Docker as well.

Instructions to run the program:

Unzip the file :

$ unzip Twitter-Akka.zip
Go to home directory of Twitter-Akka
- $ cd Twitter-Akka

For distributed client-server EXECUTION,

First step, Start the server machine, * $ cd TwitterServer * $ vi src/main/resources/application.conf In this file application.conf, replace <IP_ADDRESS_OF_SERVER_MACHINE>” with ip address of the host and run the following command: * $ sbt "run <No. of Computational Clusters>" * Here, you can mention the number of computatonal clusters, each of which consists of a server and 10 workers * for example:- * $ sbt “run 6”

Second step, Start the client machine, * $ cd TwitterClient * $ vi src/main/resources/application.conf In this file application.conf, replace <IP_ADDRESS_OF_CLIENT_MACHINE>” with ip address of the client and run the following command:

 $ sbt “run <IP_ADDRESS_OF_SERVER_MACHINE>”

```
 for example:-
```
```
 $ sbt "run 128.227.170.242"
```

To stop

To stop server, press CTRL+C

TO RUN ON THE LINUX MACHINES PROVIDED:

Use putty and type pcluster.cise.ufl.edu and port number 2222 for machine p2, the ip for this is 192.168.1.2
Enter your gatorlink as username and gatorid as password
Use 'git clone' to clone the Twitter repository into the machine
You can access this folder through WinSCP to create and add common jar files to the folders
Change the application.conf files of Server, Client and RestAPI servers to include ip address 192.168.1.2
Change the MyServer.scala file in RestAPI folder, to replace 'localhost' with 192.168.1.2 and port number 8080 with 8081(optional)
Start the Server and RestAPIServer on p2
sbt "run 5" on server and sbt "run 192.168.1.2" on RestAPIServer
In the same way, start another machine p3 from putty, using 2322 as port number, the ip for this is 192.168.1.3
From this machine p3, we can run the curl commands to access p2 machine such as:
curl "http://192.168.1.2/8081/serach?searchString=Hello"

soumitraj/Twitter-Akka

Twitter-Akka

Instructions to run the program: