packet loss in larger packets

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

packet loss in larger packets

Erwin Lubbers-2
Hi,

I'm using OpenBSD 5.1 and an Intel 10GbE SR (82598AF) ethernet card as a
router/firewall and it's working almost perfect. It is routing around 2 gbps
of traffic.

On the ix0 interface there are several vlans configured with an MTU of 1500.
When I'm pinging a switch connected to the system (with 10 gbps) there is no
packet loss while sending packets of 1472 bytes. From 1473 bytes and more
there is somewhere between 15 to 40% loss.

I first thought the switch was busy, but from another (Linux) system,
connected with 10 gbps and the same network interface there is no loss on
larger packets.

Does someone have an idea on how to solve this?

Regards,
Erwin

Reply | Threaded
Open this post in threaded view
|

Re: packet loss in larger packets

Camiel Dobbelaar
On Fri, 21 Sep 2012, Erwin Lubbers wrote:

> I'm using OpenBSD 5.1 and an Intel 10GbE SR (82598AF) ethernet card as a
> router/firewall and it's working almost perfect. It is routing around 2 gbps
> of traffic.
>
> On the ix0 interface there are several vlans configured with an MTU of 1500.
> When I'm pinging a switch connected to the system (with 10 gbps) there is no
> packet loss while sending packets of 1472 bytes. From 1473 bytes and more
> there is somewhere between 15 to 40% loss.
>
> I first thought the switch was busy, but from another (Linux) system,
> connected with 10 gbps and the same network interface there is no loss on
> larger packets.
>
> Does someone have an idea on how to solve this?

Can you show from both systems with tcpdump what the packets look like?

You are using normal (no flood) ping and the systems and switch are not
loaded with other traffic?

Reply | Threaded
Open this post in threaded view
|

Re: packet loss in larger packets

Erwin Lubbers-2
Op 21 sep. 2012, om 09:43 heeft Camiel Dobbelaar <[hidden email]> het volgende
geschreven:

>
> Can you show from both systems with tcpdump what the packets look like?
>
> You are using normal (no flood) ping and the systems and switch are not
> loaded with other traffic?
>
>


No flooding ping, just normal ping packets. I will create a tcpdump later. But
the output of a 1472 and 1473 ping packet looks like this. And even if I
disable PF the problem stays the same. Switch is handling around 350 mbps of
traffic at the moment of doing this pings.

# ping -s 1472 -c 20 10.0.1.239
PING 10.0.1.239 (10.0.1.239): 1472 data bytes
1480 bytes from 10.0.1.239: icmp_seq=0 ttl=255 time=1.782 ms
1480 bytes from 10.0.1.239: icmp_seq=1 ttl=255 time=1.499 ms
1480 bytes from 10.0.1.239: icmp_seq=2 ttl=255 time=1.244 ms
1480 bytes from 10.0.1.239: icmp_seq=3 ttl=255 time=1.339 ms
1480 bytes from 10.0.1.239: icmp_seq=4 ttl=255 time=1.453 ms
1480 bytes from 10.0.1.239: icmp_seq=5 ttl=255 time=1.486 ms
1480 bytes from 10.0.1.239: icmp_seq=6 ttl=255 time=1.627 ms
1480 bytes from 10.0.1.239: icmp_seq=7 ttl=255 time=2.323 ms
1480 bytes from 10.0.1.239: icmp_seq=8 ttl=255 time=1.386 ms
1480 bytes from 10.0.1.239: icmp_seq=9 ttl=255 time=1.511 ms
1480 bytes from 10.0.1.239: icmp_seq=10 ttl=255 time=1.578 ms
1480 bytes from 10.0.1.239: icmp_seq=11 ttl=255 time=1.552 ms
1480 bytes from 10.0.1.239: icmp_seq=12 ttl=255 time=1.732 ms
1480 bytes from 10.0.1.239: icmp_seq=13 ttl=255 time=1.279 ms
1480 bytes from 10.0.1.239: icmp_seq=14 ttl=255 time=1.369 ms
1480 bytes from 10.0.1.239: icmp_seq=15 ttl=255 time=1.399 ms
1480 bytes from 10.0.1.239: icmp_seq=16 ttl=255 time=1.513 ms
1480 bytes from 10.0.1.239: icmp_seq=17 ttl=255 time=1.546 ms
1480 bytes from 10.0.1.239: icmp_seq=18 ttl=255 time=1.551 ms
1480 bytes from 10.0.1.239: icmp_seq=19 ttl=255 time=1.483 ms
--- 10.0.1.239 ping statistics ---
20 packets transmitted, 20 packets received, 0.0% packet loss
round-trip min/avg/max/std-dev = 1.244/1.532/2.323/0.227 ms


# ping -s 1473 -c 20 10.0.1.239
PING 10.0.1.239 (10.0.1.239): 1473 data bytes
1481 bytes from 10.0.1.239: icmp_seq=1 ttl=255 time=2.107 ms
1481 bytes from 10.0.1.239: icmp_seq=2 ttl=255 time=2.035 ms
1481 bytes from 10.0.1.239: icmp_seq=3 ttl=255 time=2.045 ms
1481 bytes from 10.0.1.239: icmp_seq=4 ttl=255 time=2.048 ms
1481 bytes from 10.0.1.239: icmp_seq=6 ttl=255 time=2.708 ms
1481 bytes from 10.0.1.239: icmp_seq=7 ttl=255 time=1.768 ms
1481 bytes from 10.0.1.239: icmp_seq=8 ttl=255 time=2.274 ms
1481 bytes from 10.0.1.239: icmp_seq=9 ttl=255 time=1.775 ms
1481 bytes from 10.0.1.239: icmp_seq=11 ttl=255 time=3.969 ms
1481 bytes from 10.0.1.239: icmp_seq=13 ttl=255 time=5.679 ms
1481 bytes from 10.0.1.239: icmp_seq=14 ttl=255 time=2.012 ms
1481 bytes from 10.0.1.239: icmp_seq=15 ttl=255 time=2.148 ms
1481 bytes from 10.0.1.239: icmp_seq=17 ttl=255 time=2.179 ms
1481 bytes from 10.0.1.239: icmp_seq=18 ttl=255 time=1.796 ms
1481 bytes from 10.0.1.239: icmp_seq=19 ttl=255 time=3.433 ms
--- 10.0.1.239 ping statistics ---
20 packets transmitted, 15 packets received, 25.0% packet loss
round-trip min/avg/max/std-dev = 1.768/2.531/5.679/1.035 ms