Github user joey commented on the pull request:
https://github.com/apache/incubator-nifi/pull/51#issuecomment-120740843
Sorry, hit comment by accident.
This matches the behavior of Flume so there's no loss in functionality for
the user. In fact a number of sinks/sources have config parameters to run
threads internally for parallelism so users can still take advantage of those.
Finally, I reviewed some Flume code and I'm nervous about finding subtle
concurrency issues with trying to run multi-threaded sinks/sources.
If performance is a real concern as users start working with it, we can
work on adding it in the future.
I should have an updated PR with the other comments addressed later today.
As far as a drop down of sources/sinks go, Flume has no service loader
mechanism so I can't do it generally. I could populate a combo box with the
sources/sinks we ship and users could type in a class if they added one after
the fact.
There is no API for introspecting config parameters so we'll have to keep
that a copy/paste experience. I agree with Ryan, if any sources/sinks see a lot
of adoption we should re-write them as first-class processors.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
