These exemplary Consumers can processes live Netflow data as generated by the bwNetFlow platform. Usually, this is flow data from border interfaces, possible filtered to a customer-specific level of granularity.

For most examples, you will need an authdata file (it's in .gitignore) which needs to look like this:

username-groupname
username
password

Note that the groupname is enforced to have your username as a prefix for namespacing, the rest may be descriptive or random. None of the examples make use of paralellism or use distinc groupnames, so just one example should be run at a time, or else Kafka will split the flows between the examples, as they are sharing a group.

Additionally, anon_demo.py works as is, without any authdata. Your user should of course have the proper access level for the topic specified inside the script. It usually is set to flow-messages-enriched, which users at a participant network will need to change to their custom topic.

The local virtualenv needs to contain the packages confluent-kafka and protobuf.