Hi Stanislav,
punt-dispatch should be doing it; below is an example of packet trace from my
test env:
...
00:00:11:655232: ip4-receive
UDP: 192.168.10.1 -> 10.0.0.2
tos 0x00, ttl 63, length 280, checksum 0xc2c9 dscp CS0 ecn NON_ECN
fragment id 0xa360, flags DONT_FRAGMENT
Is it an error ?
vl_api_sockclnt_delete_t_handler():
u32 reg_index = socket_api_registration_handle_to_index (ntohl (mp->index));
ntohl(mp->index) seems like to be ntohl(mp->client_index)
otherwise , it can't work on multiple clients.
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages
Hi Filip,
In my case an initiator is behind NAT while a responder has a public IP.
What node should check for SPI=0? I also have LCP enabled on some
interfaces, maybe it breaks the default punt behaviour.
On Fri, 13 May 2022 at 10:54, Filip Tehlar -X (ftehlar - PANTHEON TECH SRO
at Cisco)
I'm not entirely sure that's the case - the reason being that default port 4500
is used for both nat traversal and encrypted dataplane traffic. The way to
distinguish between those two is having SPI=0 in case of NATT. For this there
is punt mechanism in vpp, which forwards 4500 IKE packets to
Hmm good catch, I wonder why we did not catch it - maybe the unit tests use
only non-standard port...
Can you try to patch it accordingly and if it solves the issue, push it on
gerrit for review?
If you can update the unit tests to reproduce the issue and verify the fix
fixes it it would be