Solved LACP Hash Policy

S

shahzaib

Hi,

We've configured 2x1Gbps LACP on FreeBSD recently and observed that traffic is always transmitting via single interface while the other interface is always idle. So as Linux has bond_xmit_hash_policy layer2+3 to load-balance traffic between both interfaces, does FreeBSD supports this as well ? Here are the LACP config of server :

Code: 
ifconfig_igb0="up"
ifconfig_igb2="up"
cloned_interfaces="lagg0"
ifconfig_lagg0="laggproto lacp laggport igb0 laggport igb2 1.1.1.1 netmask 255.255.255.224"
defaultrouter="1.1.1.10"

I've read that there's option such as lagghash l3,l4 to do that, so should I put it as :

Code: 
ifconfig_lagg0="laggproto lacp laggport igb0 laggport igb2 1.1.1.1 netmask 255.255.255.224 lagghash l3,l4"

Thanks in advance!
 
Put all the lagg* options together, in a group, in front of the IP address info:
Code: 
ifconfig_lagg0="laggproto lacp laggport igb0 laggport igb2 lagghash l3,l4 inet 1.1.1.1 netmask 255.255.255.224"

It's not strictly required, but it streamlines things, keeps things grouped according to function, and some ifconfig options reset an interface causing some bits to be missed by the RC system. Alternately, put the IP info first, with all the non-IP information afterward:

Code: 
ifconfig_lagg0="inet 1.1.1.1 netmask 255.255.255.224 laggproto lacp laggport igb0 laggport igb2 lagghash l3,l4"
 
Hi, it still didnt fixed the issue. Inward traffic is providing full 2Gbps throughput (Both interface give full thrust to download) but outgoing traffic is stick to single interface while 2nd interface only gives 15Mbps at peaks and throughput doesn't exceed 800-900Mbps. Here is my config now :

Code: 
ifconfig_lagg0="inet 1.1.1.1 netmask 255.255.255.224 laggproto lacp laggport igb0 laggport igb2 lagghash l3,l4"
 
It came out to be issue with kernel configs. We've fixed it and traffic is properly load-balanced on 2Gbps Full-Duplex.

Code: 
sysctl net.link.lagg.0.use_flowid=0
 
You must log in or register to reply here.
Back
Top