![Metrodata NT10G05 Скачать руководство пользователя страница 64](http://html1.mh-extra.com/html/metrodata/nt10g05/nt10g05_user-manual_1779477064.webp)
NT10G05 USER MANUAL
76-02-105 Rev.B
Page 64 of 80
4.2.3.2.1.7.1
Perform Test
To perform the RFC-2544 throughput test, perform test should be set to yes. Set to
no to disable. Note, since the latency test relies on the throughput test it will also be
disabled if throughput is disabled.
4.2.3.2.1.7.2
Short/Long Test Time
This defines the duration of the hunt and extended throughput test phases. The
minimum setting is 10 seconds.
This test duration must be set to give confidence in the network performance, though
the default setting is a good compromise between trust and overall test time.
If it turns out that the 10 second hunt window didn’t lose packets due to buffering it is
likely that the performance test will drop packets, in this case the bandwidth will back
off 1% and the test will repeat.
4.2.3.2.1.7.3
Upper Limit
Normally, RFC-2544 is a test of the physical port and commences at wire speed. In
some cases however, it is undesirable to fully utilise the link and so a lower starting
rate may be defined. For example, an NT10G05 with a 10Gigabit uplink is testing
across a 100M circuit. In this case the max rate may be defined to be 100M.
4.2.3.2.1.8 Latency Test (L2/L3)
Once the throughput has been discovered for each packet size, a latency test is
performed where every packet is checked for the latency and inter packet delay
variation.