Fetching aggregate statistics would be better. I believe your problem of
negative count happens when some flows expire and thus are removed from the
table. Other option would be to remember both old and new flow stats and
count only ones which did not expire inbetween
Peter
(sent from android devi
Hi,
I'm trying to fetch some statistics such as bytes_count of a certain switch.
I use *ofp_flow_stats_request()* to send the request and handle it this way:
for f in event.stats : bytes += f.byte_count
In order to get the byte_count of a certain interval, I send it
periodically and use *current
