Any reason not to just use one of the existing MongoDB stages? What's your use case?

Use Case : i should drop the existing collection before loading the new collection.I though of using groovy script to handle in mongodb (drop & create collection) it is truncate & load scenario.

If you want to do this once per pipeline run, I'd simply use the Pipeline Start event to call a Shell executor and have the mongo shell drop your collection for you (you'd just need to have that installed on your Data Collector node). Seems a lot easier than dealing with packages and Groovy code.