Files
ferrix/crates/integration-tests/tests/negotiation.rs
T
Rayhaan JaufeerallyandClaude Opus 5.5 65f136048e Graceful restart (RFC 4724, RFC 8538)
A peer that restarts keeps its routes with us, and we keep ours with our
peers while we restart, so that a restart withdraws nothing across the
network and the machine goes on forwarding. On by default for every
peer; `graceful_restart` sets the restart time we give, caps how long a
peer's routes are kept, or turns it off.

When a peer restarts: its routes are marked stale in the routing table,
and chosen as before, for at most its restart time. When it is back, what
it sends again is fresh, and the stale routes of a family go at its
End-of-RIB marker, or after the cap; at once for a family whose
forwarding it did not keep. Only a connection that failed counts as a
restart, and, with RFC 8538 agreed, an expired hold timer or a
NOTIFICATION other than a Hard Reset. The stale mark is a bit of the
route's word that selection never looks at.

When we restart: SIGUSR1 leaves the kernel's routes where they are,
closes the sessions without a word where graceful restart is agreed, and
exits with status 75, on which the systemd unit starts us again. The next
start finds the routes in the kernel, says so in its OPENs (R, and F where
forwarding was kept), and has them go once every peer has sent its
End-of-RIB. Our own End-of-RIB markers for learned routes wait for that
too. SIGTERM still stops for good, now with a Hard Reset where RFC 8538 is
agreed, as are errors of ours.

The status page and API show what was agreed with each peer, the routes
kept while a peer restarts, and which routes are stale. Tested against
FRR and BIRD in a new phase of the interoperability lab, and with routes
kept in a real kernel over a restart.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-04 22:43:47 +02:00

328 lines
11 KiB
Rust

//! End-to-end tests for capability negotiation: what the two OPENs agree on
//! decides which routes are sent and accepted, and how they are encoded.
use std::net::{Ipv4Addr, Ipv6Addr, SocketAddr};
use std::time::Duration;
use bgp_integration_tests::test_peer::TestPeer;
use bgp_integration_tests::test_server::TestServer;
use bgp_packet::errors::{BgpError, OpenMsgSubcode};
use bgp_packet::message::{
AsPathAttribute, Message, MpReachNlriPathAttribute, NextHopPathAttribute, NlriNextHop,
OriginPathAttribute, PathAttribute, UpdateMessage,
};
use bgp_packet::open::Capability;
use bgp_primitives::constants::{Afi, AfiSafi, Safi};
use bgp_primitives::ip_prefix::IpPrefix;
use bgp_rib::{LocalRoute, LocalRouteAttrs};
use bgp_server::config::PeerConfig;
use bgp_server::family::{FamilySet, SUPPORTED};
use bgp_server::policy::ImportPolicy;
use eyre::{Result, bail};
const LOCAL_ASN: u32 = 65000;
const PEER_ASN: u32 = 65001;
const PEER_ID: Ipv4Addr = Ipv4Addr::new(10, 0, 0, 1);
const TIMEOUT: Duration = Duration::from_secs(5);
fn prefix(s: &str) -> IpPrefix {
IpPrefix::try_from(s).unwrap()
}
fn only(family: AfiSafi) -> FamilySet {
[family].into_iter().collect()
}
fn four_octet() -> Capability {
Capability::FourByteAsn { asn: PEER_ASN }
}
async fn wait_for_session(server: &TestServer) -> Result<bgp_server::session::SessionHandle> {
Ok(tokio::time::timeout(TIMEOUT, async {
loop {
if let Some(session) = server.registry.list().into_iter().next() {
return session;
}
tokio::time::sleep(Duration::from_millis(10)).await;
}
})
.await?)
}
/// An UPDATE announcing one IPv4 route in the body and one IPv6 route in
/// MP_REACH_NLRI.
fn dual_stack_update() -> UpdateMessage {
UpdateMessage {
withdrawn_routes: vec![],
path_attributes: vec![
PathAttribute::Origin(OriginPathAttribute::IGP),
AsPathAttribute::from_asns(vec![PEER_ASN]),
PathAttribute::NextHop(NextHopPathAttribute(Ipv4Addr::new(192, 0, 2, 1))),
PathAttribute::MpReachNlri(MpReachNlriPathAttribute {
family: AfiSafi::IPV6_UNICAST,
next_hop: NlriNextHop::Ipv6(Ipv6Addr::new(0x2001, 0xdb8, 0, 0, 0, 0, 0, 1)),
prefixes: vec![prefix("2001:db8:1::/48").into()],
}),
],
announced_routes: vec![prefix("192.0.2.0/24").into()],
}
}
/// The session carries the families both OPENs offered, and the server's
/// OPEN offers every family by default.
#[tokio::test]
async fn test_families_are_the_ones_both_sides_offered() -> Result<()> {
let server = TestServer::start(LOCAL_ASN).await?;
let mut peer = TestPeer::connect(server.addr).await?;
let open = peer
.handshake_with(
PEER_ASN,
PEER_ID,
90,
vec![
four_octet(),
Capability::multiprotocol(AfiSafi::IPV4_UNICAST),
],
)
.await?;
for family in SUPPORTED {
assert!(
open.capabilities()
.any(|c| *c == Capability::multiprotocol(family)),
"the server's OPEN offers {family}"
);
}
let session = wait_for_session(&server).await?;
assert_eq!(session.families, only(AfiSafi::IPV4_UNICAST));
Ok(())
}
/// A route of a family the peer did not offer is not accepted from it. The
/// rest of the UPDATE is, and the session stays up.
#[tokio::test]
async fn test_routes_of_a_family_not_negotiated_are_not_accepted() -> Result<()> {
let server = TestServer::start(LOCAL_ASN).await?;
let mut peer = TestPeer::connect(server.addr).await?;
peer.handshake_with(
PEER_ASN,
PEER_ID,
90,
vec![
four_octet(),
Capability::multiprotocol(AfiSafi::IPV4_UNICAST),
],
)
.await?;
wait_for_session(&server).await?;
peer.send(&Message::Update(dual_stack_update())).await?;
let routes = tokio::time::timeout(TIMEOUT, async {
loop {
let routes = server.received_routes();
if !routes.is_empty() {
return routes;
}
tokio::time::sleep(Duration::from_millis(10)).await;
}
})
.await?;
// One UPDATE is applied as a whole, so the IPv6 route is not on its way.
assert_eq!(routes.len(), 1);
assert_eq!(routes[0].prefix, prefix("192.0.2.0/24"));
assert_eq!(server.registry.list().len(), 1, "the session survives");
Ok(())
}
/// A route of a family the peer did not offer is not sent to it.
#[tokio::test]
async fn test_routes_of_a_family_not_negotiated_are_not_sent() -> Result<()> {
let server = TestServer::start_with(LOCAL_ASN, |c| {
c.originate = vec![LocalRoute::new(
prefix("10.100.0.0/16"),
LocalRouteAttrs::default(),
)];
})
.await?;
// This peer wants IPv6 only; the server has an IPv4 route.
let mut v6_only = TestPeer::connect(server.addr).await?;
v6_only
.handshake_with(
PEER_ASN,
PEER_ID,
90,
vec![
four_octet(),
Capability::multiprotocol(AfiSafi::IPV6_UNICAST),
],
)
.await?;
let session = wait_for_session(&server).await?;
assert_eq!(session.families, only(AfiSafi::IPV6_UNICAST));
// A second peer that takes IPv4 is sent the route. The advertiser
// handles sessions in the order they came up, so by the time this one
// has its route the first has been dealt with.
let mut dual = TestPeer::connect(server.addr).await?;
dual.handshake(PEER_ASN, Ipv4Addr::new(10, 0, 0, 2), 90)
.await?;
let update = tokio::time::timeout(TIMEOUT, dual.recv_update()).await??;
assert_eq!(update.announced_routes, vec![prefix("10.100.0.0/16")]);
let nothing = tokio::time::timeout(Duration::from_millis(200), v6_only.recv_update()).await;
assert!(nothing.is_err(), "got {nothing:?}");
Ok(())
}
/// A configured peer is offered only the families configured for it, and a
/// peer that shares none of them is refused with the capability we needed.
#[tokio::test]
async fn test_no_family_in_common_refuses_the_session() -> Result<()> {
let server = TestServer::start_with(LOCAL_ASN, |c| {
// Inbound connections from this address inherit the peer's settings.
c.peers = vec![PeerConfig {
graceful_restart: Default::default(),
name: "v6-only".into(),
endpoint: SocketAddr::new(Ipv4Addr::LOCALHOST.into(), 1).into(),
remote_asn: PEER_ASN,
hold_time_secs: 90,
local_address: None,
families: only(AfiSafi::IPV6_UNICAST),
add_path: None,
import: ImportPolicy::permissive(),
role: None,
require_role: false,
export: Default::default(),
}];
// Keep the dialer out of the way.
c.connect_retry = Duration::from_secs(3600);
})
.await?;
let mut peer = TestPeer::connect(server.addr).await?;
peer.send_open_with(
PEER_ASN,
PEER_ID,
90,
vec![
four_octet(),
Capability::multiprotocol(AfiSafi::IPV4_UNICAST),
],
)
.await?;
let open = peer.recv_open().await?;
let offered: Vec<_> = open
.capabilities()
.filter(|c| matches!(c, Capability::MultiProtocol { .. }))
.collect();
assert_eq!(offered, [&Capability::multiprotocol(AfiSafi::IPV6_UNICAST)]);
match tokio::time::timeout(TIMEOUT, peer.recv()).await?? {
Message::Notification(n) => {
assert_eq!(
n.error,
BgpError::OpenMsg(OpenMsgSubcode::UnsupportedCapability)
);
// The Multiprotocol capability for IPv6 unicast.
assert_eq!(n.data, [1, 4, 0, 2, 0, 1]);
}
other => bail!("expected NOTIFICATION, got {other:?}"),
}
assert!(server.registry.list().is_empty());
Ok(())
}
/// Real routers offer families we do not support. That must not stop
/// the families we do share from working.
#[tokio::test]
async fn test_a_peer_offering_an_unsupported_family_is_still_accepted() -> Result<()> {
let server = TestServer::start(LOCAL_ASN).await?;
let mut peer = TestPeer::connect(server.addr).await?;
peer.handshake_with(
PEER_ASN,
PEER_ID,
90,
vec![
four_octet(),
Capability::multiprotocol(AfiSafi::new(Afi::L2VPN, Safi::EVPN)),
Capability::multiprotocol(AfiSafi::IPV4_UNICAST),
],
)
.await?;
let session = wait_for_session(&server).await?;
assert_eq!(session.families, only(AfiSafi::IPV4_UNICAST));
assert!(
session.four_octet_asn,
"capabilities after it are still read"
);
Ok(())
}
/// A peer that offers no capabilities at all is a plain RFC 4271 speaker:
/// IPv4 unicast, two-octet AS numbers. Our four-octet AS number reaches it
/// as AS_TRANS in AS_PATH and in full in AS4_PATH (RFC 6793 §4.2.2).
#[tokio::test]
async fn test_a_peer_without_capabilities_gets_two_octet_as_paths() -> Result<()> {
const WIDE_ASN: u32 = 4_200_000_001;
let server = TestServer::start_with(WIDE_ASN, |c| {
c.originate = vec![LocalRoute::new(
prefix("10.100.0.0/16"),
LocalRouteAttrs::default(),
)];
})
.await?;
let mut peer = TestPeer::connect(server.addr).await?;
peer.handshake_with(PEER_ASN, PEER_ID, 90, vec![]).await?;
let session = wait_for_session(&server).await?;
assert!(!session.four_octet_asn);
assert_eq!(session.families, only(AfiSafi::IPV4_UNICAST));
// The test peer now decodes AS_PATH two octets wide, as such a speaker
// would; an UPDATE encoded four octets wide would not parse.
let update = tokio::time::timeout(TIMEOUT, peer.recv_update()).await??;
assert_eq!(update.announced_routes, vec![prefix("10.100.0.0/16")]);
assert!(
update
.path_attributes
.contains(&AsPathAttribute::from_asns(vec![23456])),
"{update:?}"
);
let as4_path = update.path_attributes.iter().find_map(|a| match a {
PathAttribute::As4Path(path) => Some(path.segments[0].path.clone()),
_ => None,
});
assert_eq!(as4_path, Some(vec![WIDE_ASN]));
// And its own two-octet paths are read correctly.
peer.send(&Message::Update(UpdateMessage {
withdrawn_routes: vec![],
path_attributes: vec![
PathAttribute::Origin(OriginPathAttribute::IGP),
AsPathAttribute::from_asns(vec![PEER_ASN, 64512]),
PathAttribute::NextHop(NextHopPathAttribute(Ipv4Addr::new(192, 0, 2, 1))),
],
announced_routes: vec![prefix("192.0.2.0/24").into()],
}))
.await?;
let routes = tokio::time::timeout(TIMEOUT, async {
loop {
let routes = server.received_routes();
if !routes.is_empty() {
return routes;
}
tokio::time::sleep(Duration::from_millis(10)).await;
}
})
.await?;
assert!(
routes[0]
.attributes
.contains(&AsPathAttribute::from_asns(vec![PEER_ASN, 64512]))
);
Ok(())
}