When to use Flink
Select Flink as your data processing technology when:
- You need high performance. Flink at the moment is one of the best in performance for stream processing.
- Your use case needs machine learning. Flink’s native closed loop iterations operators make the processing perform much faster.
- Your use case needs graph processing. Again, because of the preceding same feature, Flink will process data faster.
- You require high throughput rates with guaranteed consistency.
- You need exactly one time processing. This also eliminates duplicate record processing.
- You want to avoid handling memory manually and leave that to the framework. Flink has automatic memory management.
- You need to deal with intermediate results and Flink follows the data flow approach making it easy to do this.
- You need less configuration. Many aspects in Flink are abstracted away from the user and this makes configuration simple.
- You need to deal with both batch and stream data using the same framework. Flink is a hybrid...
