BGP: Allow multiple EBGP neighbors with the same peer.
We can distinguish BGP sessions if at least one side uses a different IP
address. Extend olock mechanism to handle local IP as a part of key, with
optional wildcard, so BGP sessions could local IP in the olock and not
block themselves.
Inrin [Tue, 6 Feb 2024 18:03:14 +0000 (19:03 +0100)]
birdc: Do not execute cmd on noninteractive help request
The help command triggered by '?' keeps the message in readline buffer,
so it could be edited. For noninteractive shell it leads to an unexpected
side effect that `echo <cmd> ? | birdc` executes the command <cmd> after
showing its help. Avoid this by clearing the readline buffer in such case.
Ondrej Zajicek [Sat, 27 Jan 2024 16:38:06 +0000 (17:38 +0100)]
Nest: Fix bug in recursive routes with MPLS-labeled nexthops
When a recursive route with MPLS-labeled nexthop was exported to kernel
and read back, the nexthop_same() failed due to different labels_orig
field and kernel protocol reinstalled it unnecessarily.
For comparing hext hops, route cache has to distinguish ones with
different labels_orig, but KRT has to ignore that, so we need two
nexthop compare functions.
Implement BGP Send hold timer according to draft-ietf-idr-bgp-sendholdtimer.
The Send hold timer drops the session if the neighbor is sending keepalives,
but does not receive our messages, causing the TCP connection to stall.
Ondrej Zajicek [Thu, 23 Nov 2023 19:54:22 +0000 (20:54 +0100)]
BGP: Add options to require BGP capabilities
Some BGP capabilities change the BGP behavior in a significant way, so if
the configuration depends on it, it is better to not establish BGP
session when the capability is not available.
Add several BGP option to require individual BGP capabilities during
session negotiation.
Pavel Šorejs [Fri, 6 Oct 2023 02:31:19 +0000 (04:31 +0200)]
KRT: Allow to learn routes with RTPROT_KERNEL
The Kernel protocol, even with the option 'learn' enabled, ignores
direct routes created by the OS kernel (on Linux these are routes
with rtm_protocol == RTPROT_KERNEL).
Implement optional behavior where both OS kernel and third-party routes
are learned, it can be enabled by 'learn all' option.
Ondrej Zajicek [Thu, 5 Oct 2023 15:54:43 +0000 (17:54 +0200)]
MPLS: Handle compatibility with old configs
Old configs do not define MPLS domains and may use a static protocol
to define static MPLS routes.
When MPLS channel is the only channel of static protocol, handle it
as a main channel. Also, define implicit MPLS domain if needed and
none is defined.
Ondrej Zajicek [Wed, 4 Oct 2023 15:36:03 +0000 (17:36 +0200)]
IO: Fix race condition in event processing
When regular event was added from work event, we did remember that
regular event list was empty and therefore we did not use zero time
in poll(). This leads to ~3 s latency in route reload during
reconfiguration.
Ondrej Zajicek [Sat, 30 Sep 2023 21:18:04 +0000 (23:18 +0200)]
BGP, L3VPN: Fix MPLS channel reload
When a MPLS channel is reloaded, it should reload all regular MPLS-aware
channels. This causes re-evaluation of routes in FEC map and possibly
reannouncement of MPLS routes.
Ondrej Zajicek [Mon, 3 Oct 2022 18:06:13 +0000 (20:06 +0200)]
L3VPN: BGP/MPLS VPNs using MPLS backbone
The L3VPN protocol implements RFC 4364 BGP/MPLS VPNs using MPLS backbone.
It works similarly to pipe. It connects IP table (one per VRF) with (global)
VPN table. Routes passed from VPN table to IP table are stripped of RD and
filtered by import targets, routes passed in the other direction are extended
with RD, MPLS labels and export targets in extended communities. A separate
MPLS channel is used to announce MPLS routes for the labels.
Ondrej Zajicek [Sat, 1 Oct 2022 20:38:49 +0000 (22:38 +0200)]
MPLS: Add support for per-VRF labeling policy
The new labeling policy MPLS_POLICY_VRF assigns one label to all routes
(from the same FEC map associated with one VRF), while replaces their
next hops with a lookup to a VRF table. This is useful for L3VPN
protocol.
Ondrej Zajicek [Thu, 15 Sep 2022 00:30:15 +0000 (02:30 +0200)]
BGP: Add MPLS support
When MPLS is active, received routes on MPLS-aware SAFIs (ipvX-mpls,
vpnX-mpls) are automatically labeled according to active label policy and
corresponding MPLS routes are automatically generated. Also routes sent
on MPLS-aware SAFIs announce local labels when it should be done.
Ondrej Zajicek [Thu, 15 Sep 2022 00:29:12 +0000 (02:29 +0200)]
Static: Add MPLS support
When MPLS is active, static IP/VPN routes are automatically labeled
according to active label policy and corresponding MPLS routes are
automatically generated.
Ondrej Zajicek [Wed, 14 Sep 2022 23:38:18 +0000 (01:38 +0200)]
MPLS subsystem
The MPLS subsystem manages MPLS labels and handles their allocation to
MPLS-aware routing protocols. These labels are then attached to IP or VPN
routes representing label switched paths -- LSPs.
There was already a preliminary MPLS support consisting of MPLS label
net_addr, MPLS routing tables with static MPLS routes, remote labels in
next hops, and kernel protocol support.
This patch adds the MPLS domain as a basic structure representing local
label space with dynamic label allocator and configurable label ranges.
To represent LSPs, allocated local labels can be attached as route
attributes to IP or VPN routes with local labels as attributes.
There are several steps for handling LSP routes in routing protocols --
deciding to which forwarding equivalence class (FEC) the LSP route
belongs, allocating labels for new FECs, announcing MPLS routes for new
FECs, attaching labels to LSP routes. The FEC map structure implements
basic code for managing FECs in routing protocols, therefore existing
protocols can be made MPLS-aware by adding FEC map and delegating
most work related to local label management to it.
Ondrej Zajicek [Mon, 2 Oct 2023 13:09:30 +0000 (15:09 +0200)]
Nest: Expand rte_src.private_id to u64
In general, private_id is sparse and protocols may want to map some
internal values directly into it. For example, L3VPN needs to
map VPN route discriminators to private_id.
OTOH, u32 is enough for global_id, as these identifiers are dense.
Igor Putovny [Wed, 21 Jun 2023 11:15:07 +0000 (13:15 +0200)]
Basic route aggregation
Add a new protocol offering route aggregation.
User can specify list of route attributes in the configuration file and
run route aggregation on the export side of the pipe protocol. Routes are
sorted and for every group of equivalent routes new route is created and
exported to the routing table. It is also possible to specify filter
which will run for every route before aggregation.
Furthermore, it will be possible to set attributes of new routes
according to attributes of the aggregated routes.
This is a work in progress.
Original work by Igor Putovny, subsequent cleanups and finalization by
Maria Matejka.
Ondrej Zajicek [Wed, 13 Sep 2023 04:21:26 +0000 (06:21 +0200)]
Filter: Use common initializer for undefined variables and eattrs.
Undefined paths and clists should use typed f_val with empty adata
instead of just void f_val. Use common initializer to handle both
variables and eattrs.
Ondrej Zajicek [Wed, 12 Jul 2023 18:01:03 +0000 (20:01 +0200)]
Filter: Move argument list reversal from function_call to var_list
List of arguments for function calls is constructed in reverse and then
reverted. This was done in function_call grammar rule. Do the reverse
directly in var_list grammar rule. This fixes reverse order of arguments
in method calls.
Ondrej Zajicek [Mon, 3 Jul 2023 15:00:58 +0000 (17:00 +0200)]
Filter: Implement multiple dispatch for methods
- Extend method descriptors with type signature
- Daisy chain method descriptors for the same symbol
- Dispatch methods for same symbol based on type signature
- Split add/delete/filter operations to multiple methods
- Replace ad-hoc dispatch of old-style syntax with scope-based dispatch
- Also change method->arg_num to count initial arg
It still needs some improvements, like better handling of untyped
expressions and better error reporting when no dispatch can be done.
The multiple dispatch could also be extended to dispatch regular
function-like expressions in a uniform way.
Maria Matejka [Fri, 16 Jun 2023 15:35:37 +0000 (17:35 +0200)]
Filter: Methods rework
Methods can now be called as x.m(y), as long as x can have its type
inferred in config time. If used as a command, it modifies the object,
if used as a value, it keeps the original object intact.
Also functions add(x,y), delete(x,y), filter(x,y) and prepend(x,y) now
spit a warning and are considered deprecated.
It's also possible to call a method on a constant, see filter/test.conf
for examples like bgp_path = +empty+.prepend(1).
Inside instruction definitions (filter/f-inst.c), a METHOD_CONSTRUCTOR()
call is added, which registers the instruction as a method for the type
of its first argument. Each type has its own method symbol table and
filter parser switches between them based on the inferred type of the
object calling the method.
Also FI_CLIST_(ADD|DELETE|FILTER) instructions have been split to allow
for this method dispatch. With type inference, it's now possible.