In this case, Elasticsearch has saved the field data for bytes_received as Keyword rather than Long. It tends to default to keyword. Depending on how you are extracting the field, you can force the type to be long… if it is in the pipeline you would use the to_long() function… or you can create a custom mapping in Elasticsearch - I wrote up an article here on how I created a custom mapping and corrected the historical data. Either way you do it, the index it is saved to will continue to keep it a keyword until it is typed correctly AND you rotate the index. That’s the short.
Also of note, new questions should be in new topics… it helps to make the forum more searchable. 