Cisco Cisco IPS 4520 Sensor Weißbuch
© 2012 Cisco and/or its affiliates. All rights reserved. This document is Cisco Public Information.
Page 11 of 13
This test is solely intended to establish the latency of a device. The given frame rate does not represent the
maximum throughput.
Table 6.
Average latency for IPS 4345
IPS 4345
Frame Size (bytes)
DUT Frame Rate (Mbps)
Average Latency (µs)
256
650
148.48
512
700
143.00
1024
700
156.28
Table 7.
Average latency for IPS 4360
IPS 4360
Frame Size (bytes)
DUT Frame Rate (Mbps)
Average Latency (µs)
256
725
103.67
512
875
104.20
1024
975
130.51
Table 8.
Average latency for IPS 4510
IPS 4510
Frame Size (bytes)
DUT Frame Rate (Mbps)
Average Latency (µs)
256
800
45.21
512
1600
74.91
1024
3000
55.03
Table 9.
Average latency for IPS 4520
IPS 4520
Frame Size (bytes)
Frame Rate (Mbps)
Average Latency (µs)
256
800
39.00
512
1600
42.55
1024
3200
54.74
7. Connection Testing
7.1. Overview
The number of connections between hosts the system can track, and the rate at which those connections are
established, can be critical for certain application types - and even for the types of devices involved. While
sessions are generally the same as TCP connections, they have different meanings for different application types
and uses. This is why most network equipment focuses on connections instead of sessions.
Connection-related metrics are most critical in transactional environments where many connections are
established from a large number of hosts. Database transactions are a common example of this.
IPS vendors that understand network traffic and its relationship with connections will publish and explain these
values. Unfortunately, this is rare.