The jar must be in local directory, if you are referring to path: /user/cloudera/home looks like HDFS path.
Please put the jar containing TwitterSource class in flume lib directory.
The data will stop once the mention iteration cycle in full. You can increase the value of rollcount parameter to fetch more data and other factor is the keyword selected for analysis.
Hope this helps.
Here is the response from the faculty:
I have shared flume-sources.jar file in Module-10 folder on Google Drive.