Final answer:
Spark has several demerits including lack of support for real-time processing, the need for manual optimization, the absence of data compression techniques, and difficulty in job debugging.
Step-by-step explanation:
Demerits of using Spark
- No support for real-time processing: Spark is not designed for real-time processing, which may limit its usability for certain applications that require immediate processing and response.
- Manual optimization: Spark requires manual optimization of code and resources to achieve maximum performance, which can be a time-consuming and complex task.
- Lack of data compression techniques: Spark does not provide built-in data compression techniques, which can result in increased storage and bandwidth requirements.
- Difficulty in job debugging: Debugging Spark jobs can be challenging due to the distributed nature of the framework, making it harder to identify and fix issues.