Hello,
I'm currently running tests on security onion with both snort and suricata in order to have some data to later analyse. My issue is that, whatever I do suricata always generates considerably more alerts than snort.
I'm using tcpreplay with pcaps I could find online (most of my tests are on the first pcap in this list
http://www.netresec.com/?page=MACCDC), and I'm doing so on identical virtual machines, running security onion production where the only difference between them is the IDS engine.
After multiple tests, suricata averages around 40k total alerts, while snort averages ~27k. They both generate 168 unique alert signatures that differ on 21 signatures between the IDS's, but the signatures that are specific to a single IDS don't justify the difference in total alerts as they occur in small numbers.
I've tried with other pcap files and the difference is always there, with suricata generating considerably more.
Is there something I'm missing ? I've tried running snort with the best possible environment I can manage, and still I can't get the alerts to match up to those of suricata.