Re: Bug and configuration MPLS error?

[David Ahern <dsahern@gmail.com>]

On 4/26/17 6:40 AM, Алексей Болдырев wrote:
> 
> 
> 26.04.2017, 05:23, "David Ahern" <dsahern@gmail.com>:
>> On 4/25/17 11:28 AM, Алексей Болдырев wrote:
>>>  226 sysctl -w net.mpls.conf.lo.input=1
>>>  227 sysctl -w net.mpls.platform_labels=1048575
>>>  228 ip link add veth0 type veth peer name veth1
>>>  229 ip link add veth2 type veth peer name veth3
>>>  230 sysctl -w net.mpls.conf.veth0.input=1
>>>  231 sysctl -w net.mpls.conf.veth2.input=1
>>>  232 ifconfig veth0 10.3.3.1 netmask 255.255.255.0
>>>  233 ifconfig veth2 10.4.4.1 netmask 255.255.255.0
>>>  234 ip netns add host1
>>>  235 ip netns add host2
>>>  236 ip link set veth1 netns host1
>>>  237 ip link set veth3 netns host2
>>>  238 ip netns exec host1 ifconfig veth1 10.3.3.2 netmask 255.255.255.0 up
>>>  239 ip netns exec host2 ifconfig veth3 10.4.4.2 netmask 255.255.255.0 up
>>>  240 ip netns exec host1 ip route add 10.10.10.2/32 encap mpls 112 via inet 10.3.3.1
>>>  241 ip netns exec host2 ip route add 10.10.10.1/32 encap mpls 111 via inet 10.4.4.1
>>>  242 ip -f mpls route add 111 via inet 10.3.3.2
>>>  243 ip -f mpls route add 112 via inet 10.4.4.2
>>
>> your setup is incomplete.
>>
>> # ip netns exec host2 ping 10.10.10.1
>> PING 10.10.10.1 (10.10.10.1) 56(84) bytes of data.
>> ^C
>> --- 10.10.10.1 ping statistics ---
>> 2 packets transmitted, 0 received, 100% packet loss, time 1038ms
>>
>> If you run tcpdump on veth1 in host1 you see the packets come in but no
>> response:
>>
>> # ip netns exec host1 tcpdump -n -i veth1
>> tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
>> listening on veth1, link-type EN10MB (Ethernet), capture size 262144 bytes
>> 19:20:24.599529 IP6 fe80::347d:e3ff:fe93:944b > ff02::2: ICMP6, router
>> solicitation, length 16
>> 19:20:27.413901 IP 10.4.4.2 > 10.10.10.1: ICMP echo request, id 978, seq
>> 1, length 64
>> 19:20:28.439574 IP 10.4.4.2 > 10.10.10.1: ICMP echo request, id 978, seq
>> 2, length 64
>>
>> and the lack of response is b/c:
>> 1. host1 has no address for 10.10.10.1 and
>> 2. even if it did, there is no return route to 10.4.4.2:
>>
>> # ip -netns host1 ro ls
>> 10.3.3.0/24 dev veth1 proto kernel scope link src 10.3.3.2
>> 10.10.10.2 encap mpls 112 via 10.3.3.1 dev veth1
> 
> As for ping, you need to enter this:
> Ip netns exec host2 ping 10.10.10.1 -A 10.10.10.2
> Here I published the results of testing on new (>4.9) cores. (in Russian):
> http://forum.nag.ru/forum/index.php?s=d09f0e5186fda59b3099eb81ad07ee63&showtopic=128927
> But on the old kernels:
> http://forum.nag.ru/forum/index.php?showtopic=128927&view=findpost&p=1396067
> 

host1 does not have 10.10.10.1 as a local address.
host2 does not have 10.10.10.2 as a local address.

Given that, host1 has no business replying to a ping destined to
10.10.10.1, and host2 will not use 10.10.10.2 as a source address.

I don't have time right now to build and test on older kernels, but
based on the network config I do not see how it can work.

If you add:
  ip -netns host1 addr add dev lo 10.10.10.1/32
  ip -netns host2 addr add dev lo 10.10.10.2/32

Then it works.