Re: [RFC net-next 2/2] selftests: seg6: add selftest for SRv6 End.DT46 Behavior

From: David Ahern
Date: Tue Jun 08 2021 - 21:17:54 EST


On 6/8/21 4:40 AM, Andrea Mayer wrote:
> this selftest is designed for evaluating the new SRv6 End.DT46 Behavior
> used, in this example, for implementing IPv4/IPv6 L3 VPN use cases.
>
> Signed-off-by: Andrea Mayer <andrea.mayer@xxxxxxxxxxx>
> Signed-off-by: Paolo Lungaroni <paolo.lungaroni@xxxxxxxxxxx>
> ---
> .../selftests/net/srv6_end_dt46_l3vpn_test.sh | 573 ++++++++++++++++++
> 1 file changed, 573 insertions(+)
> create mode 100755 tools/testing/selftests/net/srv6_end_dt46_l3vpn_test.sh
>
> diff --git a/tools/testing/selftests/net/srv6_end_dt46_l3vpn_test.sh b/tools/testing/selftests/net/srv6_end_dt46_l3vpn_test.sh
> new file mode 100755
> index 000000000000..d1e156433b98
> --- /dev/null
> +++ b/tools/testing/selftests/net/srv6_end_dt46_l3vpn_test.sh
> @@ -0,0 +1,573 @@
> +#!/bin/bash
> +# SPDX-License-Identifier: GPL-2.0
> +#
> +# author: Andrea Mayer <andrea.mayer@xxxxxxxxxxx>
> +# author: Paolo Lungaroni <paolo.lungaroni@xxxxxxxxxxx>
> +
> +# This test is designed for evaluating the new SRv6 End.DT46 Behavior used for
> +# implementing IPv4/IPv6 L3 VPN use cases.
> +#
> +# The current SRv6 code in the Linux kernel only implements SRv6 End.DT4 and
> +# End.DT6 Behaviors which can be used respectively to support IPv4-in-IPv6 and
> +# IPv6-in-IPv6 VPNs. With End.DT4 and End.DT6 it is not possible to create a
> +# single SRv6 VPN tunnel to carry both IPv4 and IPv6 traffic.
> +# The SRv6 End.DT46 Behavior implementation is meant to support the
> +# decapsulation of IPv4 and IPv6 traffic coming from a single SRv6 tunnel.
> +# Therefore, the SRv6 End.DT46 Behavior in the Linux kernel greatly simplifies
> +# the setup and operations of SRv6 VPNs.
> +#
> +# Hereafter a network diagram is shown, where two different tenants (named 100
> +# and 200) offer IPv4/IPv6 L3 VPN services allowing hosts to communicate with
> +# each other across an IPv6 network.
> +#
> +# Only hosts belonging to the same tenant (and to the same VPN) can communicate
> +# with each other. Instead, the communication among hosts of different tenants
> +# is forbidden.
> +# In other words, hosts hs-t100-1 and hs-t100-2 are connected through the
> +# IPv4/IPv6 L3 VPN of tenant 100 while hs-t200-3 and hs-t200-4 are connected
> +# using the IPv4/IPv6 L3 VPN of tenant 200. Cross connection between tenant 100
> +# and tenant 200 is forbidden and thus, for example, hs-t100-1 cannot reach
> +# hs-t200-3 and vice versa.
> +#
> +# Routers rt-1 and rt-2 implement IPv4/IPv6 L3 VPN services leveraging the SRv6
> +# architecture. The key components for such VPNs are: a) SRv6 Encap behavior,
> +# b) SRv6 End.DT46 Behavior and c) VRF.
> +#
> +# To explain how an IPv4/IPv6 L3 VPN based on SRv6 works, let us briefly
> +# consider an example where, within the same domain of tenant 100, the host
> +# hs-t100-1 pings the host hs-t100-2.
> +#
> +# First of all, L2 reachability of the host hs-t100-2 is taken into account by
> +# the router rt-1 which acts as a arp/ndp proxy.
> +#
> +# When the host hs-t100-1 sends an IPv6 or IPv4 packet destined to hs-t100-2,
> +# the router rt-1 receives the packet on the internal veth-t100 interface. Such
> +# interface is enslaved to the VRF vrf-100 whose associated table contains the
> +# SRv6 Encap route for encapsulating any IPv6 or IPv4 packet in a IPv6 plus the
> +# Segment Routing Header (SRH) packet. This packet is sent through the (IPv6)
> +# core network up to the router rt-2 that receives it on veth0 interface.
> +#
> +# The rt-2 router uses the 'localsid' routing table to process incoming
> +# IPv6+SRH packets which belong to the VPN of the tenant 100. For each of these
> +# packets, the SRv6 End.DT46 Behavior removes the outer IPv6+SRH headers and
> +# performs the lookup on the vrf-100 table using the destination address of
> +# the decapsulated IPv6 or IPv4 packet. Afterwards, the packet is sent to the
> +# host hs-t100-2 through the veth-t100 interface.
> +#
> +# The ping response follows the same processing but this time the role of rt-1
> +# and rt-2 are swapped.
> +#
> +# Of course, the IPv4/IPv6 L3 VPN for tenant 200 works exactly as the IPv4/IPv6
> +# L3 VPN for tenant 100. In this case, only hosts hs-t200-3 and hs-t200-4 are
> +# able to connect with each other.
> +#
> +#
> +# +-------------------+ +-------------------+
> +# | | | |
> +# | hs-t100-1 netns | | hs-t100-2 netns |
> +# | | | |
> +# | +-------------+ | | +-------------+ |
> +# | | veth0 | | | | veth0 | |
> +# | | cafe::1/64 | | | | cafe::2/64 | |
> +# | | 10.0.0.1/24 | | | | 10.0.0.2/24 | |
> +# | +-------------+ | | +-------------+ |
> +# | . | | . |
> +# +-------------------+ +-------------------+
> +# . .
> +# . .
> +# . .
> +# +-----------------------------------+ +-----------------------------------+
> +# | . | | . |
> +# | +---------------+ | | +---------------- |
> +# | | veth-t100 | | | | veth-t100 | |
> +# | | cafe::254/64 | | | | cafe::254/64 | |
> +# | | 10.0.0.254/24 | +----------+ | | +----------+ | 10.0.0.254/24 | |
> +# | +-------+-------+ | localsid | | | | localsid | +-------+-------- |
> +# | | | table | | | | table | | |
> +# | +----+----+ +----------+ | | +----------+ +----+----+ |
> +# | | vrf-100 | | | | vrf-100 | |
> +# | +---------+ +------------+ | | +------------+ +---------+ |
> +# | | veth0 | | | | veth0 | |
> +# | | fd00::1/64 |.|...|.| fd00::2/64 | |
> +# | +---------+ +------------+ | | +------------+ +---------+ |
> +# | | vrf-200 | | | | vrf-200 | |
> +# | +----+----+ | | +----+----+ |
> +# | | | | | |
> +# | +-------+-------+ | | +-------+-------- |
> +# | | veth-t200 | | | | veth-t200 | |
> +# | | cafe::254/64 | | | | cafe::254/64 | |
> +# | | 10.0.0.254/24 | | | | 10.0.0.254/24 | |
> +# | +---------------+ rt-1 netns | | rt-2 netns +---------------- |
> +# | . | | . |
> +# +-----------------------------------+ +-----------------------------------+
> +# . .
> +# . .
> +# . .
> +# . .
> +# +-------------------+ +-------------------+
> +# | . | | . |
> +# | +-------------+ | | +-------------+ |
> +# | | veth0 | | | | veth0 | |
> +# | | cafe::3/64 | | | | cafe::4/64 | |
> +# | | 10.0.0.3/24 | | | | 10.0.0.4/24 | |
> +# | +-------------+ | | +-------------+ |
> +# | | | |
> +# | hs-t200-3 netns | | hs-t200-4 netns |
> +# | | | |
> +# +-------------------+ +-------------------+
> +#
> +#
> +# ~~~~~~~~~~~~~~~~~~~~~~~~~
> +# | Network configuration |
> +# ~~~~~~~~~~~~~~~~~~~~~~~~~
> +#
> +# rt-1: localsid table (table 90)
> +# +--------------------------------------------------+
> +# |SID |Action |
> +# +--------------------------------------------------+
> +# |fc00:21:100::6046|apply SRv6 End.DT46 vrftable 100|
> +# +--------------------------------------------------+
> +# |fc00:21:200::6046|apply SRv6 End.DT46 vrftable 200|
> +# +--------------------------------------------------+
> +#
> +# rt-1: VRF tenant 100 (table 100)
> +# +---------------------------------------------------+
> +# |host |Action |
> +# +---------------------------------------------------+
> +# |cafe::2 |apply seg6 encap segs fc00:12:100::6046|
> +# +---------------------------------------------------+
> +# |cafe::/64 |forward to dev veth-t100 |
> +# +---------------------------------------------------+
> +# |10.0.0.2 |apply seg6 encap segs fc00:12:100::6046|
> +# +---------------------------------------------------+
> +# |10.0.0.0/24|forward to dev veth-t100 |
> +# +---------------------------------------------------+
> +#
> +# rt-1: VRF tenant 200 (table 200)
> +# +---------------------------------------------------+
> +# |host |Action |
> +# +---------------------------------------------------+
> +# |cafe::4 |apply seg6 encap segs fc00:12:200::6046|
> +# +---------------------------------------------------+
> +# |cafe::/64 |forward to dev veth-t200 |
> +# +---------------------------------------------------+
> +# |10.0.0.4 |apply seg6 encap segs fc00:12:200::6046|
> +# +---------------------------------------------------+
> +# |10.0.0.0/24|forward to dev veth-t200 |
> +# +---------------------------------------------------+
> +#
> +#
> +# rt-2: localsid table (table 90)
> +# +--------------------------------------------------+
> +# |SID |Action |
> +# +--------------------------------------------------+
> +# |fc00:12:100::6046|apply SRv6 End.DT46 vrftable 100|
> +# +--------------------------------------------------+
> +# |fc00:12:200::6046|apply SRv6 End.DT46 vrftable 200|
> +# +--------------------------------------------------+
> +#
> +# rt-2: VRF tenant 100 (table 100)
> +# +---------------------------------------------------+
> +# |host |Action |
> +# +---------------------------------------------------+
> +# |cafe::1 |apply seg6 encap segs fc00:21:100::6046|
> +# +---------------------------------------------------+
> +# |cafe::/64 |forward to dev veth-t100 |
> +# +---------------------------------------------------+
> +# |10.0.0.1 |apply seg6 encap segs fc00:21:100::6046|
> +# +---------------------------------------------------+
> +# |10.0.0.0/24|forward to dev veth-t100 |
> +# +---------------------------------------------------+
> +#
> +# rt-2: VRF tenant 200 (table 200)
> +# +---------------------------------------------------+
> +# |host |Action |
> +# +---------------------------------------------------+
> +# |cafe::3 |apply seg6 encap segs fc00:21:200::6046|
> +# +---------------------------------------------------+
> +# |cafe::/64 |forward to dev veth-t200 |
> +# +---------------------------------------------------+
> +# |10.0.0.3 |apply seg6 encap segs fc00:21:200::6046|
> +# +---------------------------------------------------+
> +# |10.0.0.0/24|forward to dev veth-t200 |
> +# +---------------------------------------------------+
> +#
> +

Great description of the test; thanks for adding that documentation.

Acked-by: David Ahern <dsahern@xxxxxxxxxx>