Mixing different NIC Speeds (1Gb & 10Gb) Performance Problem Question

I did not, but I just corrected that. I reset the states, waited a few mins and so far there may be a small boost in performance. I seem to be averaging now 700Mpbs instead of 500Mbps on fast.com, however if I switch ports we get the full 900+. So better but not perfect.

ngr2001

@lnguyen

1/2/8 - My 10Gb uplink to PF LAN Nic.

I just noticed that 2 of my main clients currently at 1Gb on the switch have a lot of dropped packets. However right now I am running speedtests and I am not seeing the dropped packet counter increasing, so who know under which setup all these packets were dropped.

1Gb Win 11 Client

ngr2001

@lnguyen

Should I disable flow control on my WAN NIC as a long term production setup being I am on Cable Internet ?

lnguyen

@ngr2001 I am not sure you are catching what I have said a few times across the two threads. Ethernet Flow Control is done on L2, which is what you have been configuring. TCP Flow Control is done at L4 which seems to not function when pfSense WAN is connected to DOCSIS cable modem. If you want to use Ethernet Flow control, which IMO is ugly, then you should leave it enabled.

With my setup, I don't use ethernet flow control.

stephenw10

Mmm, try with the MTU at the default 1500 so it matches the switch and client.

200k input errors is pretty significant though.

ngr2001

@stephenw10

Just tried, same diff, speeds quickly taper down to 5-600Mbps.

I clear my stats on switch, right now at least the dropped packet count is not increasing.

stephenw10

Just to be clear you are running the switch at layer2 only right?

It's hard to see how adding the switch, or just changing the port speed, can have any effect on the tcp dircetly. It could be that is some how broken upstream such that when introducing some lower lever issue it's very badly effected though.

ngr2001

@stephenw10

I am not using any layer 3 features that I am aware of right now, thus I guess it should be in layer 2. I do have flow control enabled on the switch though.

SSH@romulus#show run
Current configuration:
!
ver 08.0.95pT213
!
stack unit 1
module 1 icx7250-48-port-management-module
module 2 icx7250-sfp-plus-8port-80g-module
stack-port 1/2/1
stack-port 1/2/3
!

vlan 1 name DEFAULT-VLAN by port
router-interface ve 1
!
!
symmetrical-flow-control enable
!
optical-monitor
optical-monitor non-ruckus-optic-enable
aaa authentication web-server default local
aaa authentication login default local
enable aaa console
hostname romulus
ip dhcp-client disable
ip dns server-address 10.0.0.1
ip route 0.0.0.0/0 10.0.0.1
!
no telnet server
!

clock timezone us Eastern
!
!
ntp
disable serve
server time.cloudflare.com
!
!
no web-management http
!
manager disable
!
!
manager port-list 987
!
!
interface ethernet 1/1/4
flow-control neg-on
!
interface ethernet 1/1/8
flow-control neg-on
!
interface ethernet 1/1/48
flow-control neg-on
!
interface ethernet 1/2/1
flow-control neg-on
!
interface ethernet 1/2/8
flow-control neg-on
!
interface ve 1
ip address 10.0.0.3 255.255.255.0
!

ip ssh idle-time 120
!
end
SSH@romulus#

ngr2001

@stephenw10

So I believe I upgraded to the Layer 3 firmware and enabled it from what I recall.

HW: Stackable ICX7250-48

UNIT 1: SL 1: ICX7250-48 48-port Management Module

  Software Package: **ICX7250_L3_SOFT_PACKAGE**
  Current License: l3-prem-8X10G
  P-ASIC  0: type B344, rev 01  Chip BCM56344_A0

stephenw10

Right. As long as it's not actually routing that shouldn't matter.

It sure seems like it's failing to negotiate flow control though. What does: ifconfig -vvm ix1 show in pfSense?

ngr2001

This post is deleted!

ngr2001

@stephenw10

When connected at 10Gb I get this:

When connected at 1Gb I get this:

lnguyen

@ngr2001 Maybe the 10GBASE-T SFP+ module you have is not passing along the 802.3x FC negotiation to the SFP+ port on the Brocade 7250