sourceId is not the hostname #2
Description
https://github.com/samsara/hydrant/blob/master/src/hydrant/flows/samsara.clj#L8
The sourceId is the partition-key
Would you want your twitter feed to be partitioned by host machine who is sending the data? I don't think so. This would break the same partition for same source rule.
sourceId is always depending on the dataset, for a twitter feed a good sourceId would be the twitter handle of the user sending the tweet. So that all tweets from a user will be processed by the same processor.
For wikipedia edits you might want to use either the page edited or the author, etc,
The code should be changed in such way that the source advertises its partition key.