Netflow Analysis : those data are wrong or I'm under attack?

Hi gents,

I'm recently on the "Elasticsearch log analysis" mood.

I've just set up a collector for my Netflow logs coming from about 20 devices across the Europe.

Everything seems fine, I have consistent data about flow directions and TCP/UDP bytes usage. But regarding others protocol (eg. ICMP; HOPOPT and so on) I'm getting those huge values under bytes (flow export related to 1 hour):

https://ibb.co/jEVuqo

eg. HOPOPT 1.6 TB of bytes and 5k flows in the last hour? Seems a bad reading or conversion, right?