Being the latest in this space (not really the latest, its origin dates back to 2008), it does try to cover many of the shortcomings its more popular competitors have within them. So in that league it does possess only a very few disadvantages as of now.
Some of the disadvantages associated with Flink can be bulleted as follows:
- Compared to competitors not ahead in popularity and community adoption at the time of writing this book
- Maturity in the industry is less
- Pipelined execution in Flink does have some limitation in regards to memory management (for long running pipelines) and fault tolerance
- Flink uses raw bytes as internal data representation, which if needed, can be hard to program
- Doesn't have matured APIs ...
