No relevant resource is found in the selected language.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Search


To have a better experience, please upgrade your IE browser.


Default link load balance mode lead MPLS traffic admeasure imbalance between eth-trunk physical links

Publication Date:  2012-11-21 Views:  145 Downloads:  0

Issue Description

Test s93 switch. there are 2 10G link eth-trunk connection between 2 s93 as p device, open MPLS function among p device. check the port traffic of 2 p device ec interconnection, found that the  traffic admeasure imbalance, one link took for more than 50%, other one is close to 0.

Abnormal ports situation is as follows:
Eth-Trunk94 up up 6.67% 32%
XGigabitEthernet4/0/3 up up 0.01% 53%
XGigabitEthernet5/0/3 up up 13% 11%

Alarm Information


Handling Process

Modify 2 P device eth-trunk load balance way to enhancing mode, add mpls protocol balance way in the templet. then MPLS data packet can be hashed by switch, decide which link under eth-trunk does mpls data packet use according to hash result. so the traffic between links balance recovered.
modify the load balance way command:
interface Eth-Trunk94
mode lacp-static
load-balance enhanced profile core-mpls
load-balance-profile core-mpls
mpls field top-label

the interface situation after modifying.
Eth-Trunk94 up up 7.78% 2.59%
XGigabitEthernet1/0/2 up up 7.55% 2.93%
XGigabitEthernet1/0/3 up up 8.01% 2.24%

Root Cause

The data that transfer between 2 P devcies is MPLS packet, but the load balance mode of eth-trunk member links is defaultly by traffic admeasure base on hash value that exclusive OR source IP and destination IP address of IP data packet, so the defalt load balance mode cannot balance the traffic between links.


If the traffic among eth-trunk is non-IPv4 traffic, notice that modfy eth-trunk load balance mode. load balance enhancing mode includes: 2 field, ipv4 field, ipv6 field, mpls field.