Customer has L2TP packet loss. The USG is used as LNS and clients are connecting to USG. Clients are connecting thru 3g/4g from mobile operators, but during tests it appeared that FTTB clients are also
experiencing this issue. Looks like the issue is caused by clients increase.
When it was 250 clients, it was ok, but increasing the clients led to the packet loss increase.
when ping an ip address, packet loss in tunnels increase.
PING 10.21.2.1 (10.21.2.1): 56 data bytes
Request timeout for icmp_seq 0
Request timeout for icmp_seq 1
64 bytes from 10.21.2.1: icmp_seq=2 ttl=254 time=7.543 ms
Request timeout for icmp_seq 3
64 bytes from 10.21.2.1: icmp_seq=4 ttl=254 time=7.803 ms
Request timeout for icmp_seq 5
64 bytes from 10.21.2.1: icmp_seq=6 ttl=254 time=7.550 ms
Request timeout for icmp_seq 7
64 bytes from 10.21.2.1: icmp_seq=8 ttl=254 time=7.243 ms
64 bytes from 10.21.2.1: icmp_seq=9 ttl=254 time=8.049 ms
64 bytes from 10.21.2.1: icmp_seq=10 ttl=254 time=8.703 ms
Request timeout for icmp_seq 11
64 bytes from 10.21.2.1: icmp_seq=12 ttl=254 time=7.558 ms
64 bytes from 10.21.2.1: icmp_seq=13 ttl=254 time=8.367 ms
--- 10.21.2.1 ping statistics ---
14 packets transmitted, 8 packets received, 42.9% packet loss
round-trIP min/avg/max/stddev = 7.243/7.852/8.703/0.458 ms
1. This is a packet loss, so we configured traffic statistic, check the reason of packet loss;
2. When we configured traffic statistic and ping test again, found that the below statistic:
3.The discard detail information shown "BWM car exceed pakcets discarded", it means the bandwidth is not enough, so we check where can limit the bandwidth.
4. After check the configuration of interface that carry L2TP taffic, we found that there has configured bandwidth limit, the limit is 10k.
the interface configured bandwidth limit
After modify the value of the bandwidth limit, the issue is solved.
Using the taffic statistic feature check where is the the packet loos, it can help locating the root cause.