Create a Post
cancel
Showing results for 
Search instead for 
Did you mean: 
T_L
Contributor

MHO-175 Uplink Interface Speeds

Good Afternoon - I have a weird one for the group.

We are seeing something funky with our MHO-175 uplink ports where they are displaying 10G speeds regardless of any setting setting modifications we make.

Setup:

Single Site - 2 x MHO-175 R81.20 - JHF53 on orchestrators and all SGMs

MHO1_physical port-3 (1/3/1)  ---------  MHO2-175_physical port 3 (2/3/1)

40G - CheckPoint SR optics (using MTP OM4 fiber)

Maestro QSFP port settings 40G, admin up

Orch_stat shows  40GSR optics, plugged, up/up - we have link and traffic

*We have these two uplink ports 1/3/1-2/3/1 in an 802.3ad bond (LACP) - bond1

LAN side is using 40G-SR optics and showing 40/40--80G

The issue we are seeing is the connected Security Group  is showing us 10/10--20G when we look at bond1. SNMP is showing us 10G/10G--20G.

The LAN side does seem to have any issues - we moved the LAN side ports to other interfaces, different switch / router platforms, etc - it always shows 40/40--80 and comes up -- the CP side only shows the 10/10--20.

We tried new optics on the CP side, tried going only with CP optics on both sides, tried fiber and DAC cabling - no change - 10G/10G--20G on the CP side.

The security group is utilizing 40G downlinks - and they are all showing 40.

So then we tried just testing on an empty MHO uplink interface without any bond/ LAG in the mix - enabled the Maestro port at 40G, SR4 - presented it to the Security Group in the Orchestrator - look at the interface from the SecGrp and it shows 'speed N/A' - our SNMP monitoring tools show 10G.

There are no configuration options relating to speed, duplex, negotiation for the interface when looking at it on the sec grp - so we cannot manually do anything from the SGMs.

And we also noticed that enabling/ disabling auto-negotiation on the Maestro port has no effect on the SGM interface - it always shows on.

If you have made it this far and can follow along, thanks! It is driving us nuts - especially since we are leery about moving traffic to interfaces we are not confident in. Any definitive answer has also eluded our the ATAM engineer we are working with - i figured i would take a shot and see if anyone has seen this? 

Thanks!

 

 

 

 

 

0 Kudos
1 Reply
emmap
Employee
Employee

It's a cosmetic thing, the physical port speed negotiation is handled at the MHO side and the SGMs just get an interface passed on down. If you check 'orch_stat -p' on the MHOs you'll see the correct speed. The downlinks report correctly because those ports physically exist on the SGMs.

I don't know if there's anything in the works to have the speed listed correctly on the SGMs. 

0 Kudos