'Is it better to create just a single Kafka topic for streaming tweets or several depending on the subject of the tweet?
Imagine I want to create a web for monitoring the opinions of people on certain subjects, let's call these subjects A, B and C. Now, some of those tweets mention the subject A and C, so I'd like to duplicate them to classify those tweets as A and C.
What would it be better when it comes to create kafka topics for that? Creating a single topic for all the tweets in general or creating topic A, B and C to then create three different tables for each topic in S3 or similar and then keep processing them separately with spark streaming?
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|