In order to have all tcpdump output sent out to a central location
1) use the an io.writer that writes out to the network as a simple tcp stream similar to what netcat does but using pure Go.
2) once the writer can generate that stream use a gRPC client to request a connection for that stream on a Snoopy Data Endpoint that will serve as a data hub further insights engines to plug in. Each request should allow a new stream to come in and store the data with proper metadata added on top to identify its origins like Pod, interface, timestamp etc.
3) Future work may consider adding encryption and/or tunnelling to this connection/data stream.
In order to have all tcpdump output sent out to a central location
1) use the an io.writer that writes out to the network as a simple tcp stream similar to what netcat does but using pure Go. 2) once the writer can generate that stream use a gRPC client to request a connection for that stream on a Snoopy Data Endpoint that will serve as a data hub further insights engines to plug in. Each request should allow a new stream to come in and store the data with proper metadata added on top to identify its origins like Pod, interface, timestamp etc. 3) Future work may consider adding encryption and/or tunnelling to this connection/data stream.