IMIX benchmarks

[mm404]

I want to replace my home router/firewall with something running pfSense.

I have a symmetric 1gbit internet connection.

I see you list iperf and imix speeds for each of the appliances, which is great. But for the sake of comparison, I’d like to run imix on my current setup to better understand my current capabilities.
Is there any document or wiki that roughly describes how Netgate runs the imix test?

TL;DR I am not sure if I need to go with (the bit pricey) 4100 or if 2100 will be just fine for 1 Gbit.

Thank you!

[michmoor]

@mm404 The spec sheet shows that the 2100 cant do 1Gbps of IMIX (real world) traffic..

[mm404]

@michmoor I noticed that too but I most of the time, when I actually use the full 1gbit bandwidth, it's a single upload or download (ie. more similar to the iperf test).

For that reason, I wanted to benchmark my current setup to see what my current router is capable of.

[stephenw10]

You will want the 4100. You will hit the limits of the 2100 with a real 1G WAN link.

[btspce]

Did the 23.01 release change any throughput numbers compared to 22.05 (Firewall/IMIX/VPN) ? Wondering about the 2100 especially.

[stephenw10]

There has been no significant change there. I haven't seen any measurable change in local testing.

[mm404]

@stephenw10 said in IMIX benchmarks:

There has been no significant change there. I haven't seen any measurable change in local testing.

Thanks for that info. Any chance you could share little bit how is the testing done? Thank you

[stephenw10]

In my own local testing that's just using iperf3 between hosts on the WAN and LAN.

The performance testing we do is a series of tests using TRex. We wrote blog post about it a while ago: https://www.netgate.com/blog/how-fast-can-a-porsche-gt2-rs-go

[btspce]

@stephenw10 said in IMIX benchmarks:

There has been no significant change there. I haven't seen any measurable change in local testing.

Thanks Stephen. Have a nice weekend!