The json display of the version attribute is originally an
integer. It has changed, most probably mistakenly.
> {
> "vrfId": 7,
> "vrfName": "vrf1",
> "tableVersion": 3,
> "routerId": "192.0.2.1",
> "defaultLocPrf": 100,
> "localAS": 65500,
> "routes": {
> "172.31.0.1/32": {
> "prefix": "172.31.0.1/32",
> "version": "1", <--- int or string ??
Let us fix it, by using the integer display instead.
Fixes: f9f2d188e3 ("bgpd: fix 'json detail' output structure")
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
If we have (default enabled) enabled `bgp ebgp-require-policy`, then first check
it before applying the route-maps.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
If we receive an IPv6 prefix e.g.: 2001:db8:100::/64 with nextop: 0.0.0.0, and
mp_nexthop: fc00::2, we should not treat this with an invalid nexthop because
of 0.0.0.0. We MUST check for MP_REACH attribute also and decide later if we
have at least one a valid nexthop.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
Without the patch only the best path is displayed.
With the patch, display all paths including addpaths, but only for non-JSON
output to avoid breaking existing output.
E.g.:
```
munet> r2 shi vtysh -c 'sh ip bgp nei 192.168.2.3 advertised-routes'
Network Next Hop Metric LocPrf Weight Path
*> 172.16.16.254/32 192.168.2.3 0 0 65003 ?
* 172.16.16.254/32 192.168.2.4 0 0 65004 ?
*> 192.168.2.0/24 192.168.2.3 0 0 65003 ?
* 192.168.2.0/24 192.168.2.4 0 0 65004 ?
```
Before it was:
```
munet> r2 shi vtysh -c 'sh ip bgp nei 192.168.2.3 advertised-routes'
Network Next Hop Metric LocPrf Weight Path
*> 172.16.16.254/32 192.168.2.3 0 0 65003 ?
*> 192.168.2.0/24 192.168.2.3 0 0 65003 ?
```
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
A redistribute cmd can have a route-map attached to it and adding the
match source-protocol to that route-map means BGP to filter which
protocol routes to accept among the bunch of routes zebra is sending.
Fixing this since this wasnt implemented earlier.
Ticket :#4119692
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
We have this:
if ( (safi == SAFI_UNICAST) && ...)
do stuff
if ( (safi == SAFI_MPLS_VPN) && ... )
do stuff
this leads to having to test safi multiple times if safi is
SAFI_UNICAST. Let's make it a else if as that we know that
the safi is going to not change.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
In an attempt to make the code faster let's just pass
in the prefix instead of having to do a lookup a majillion
times again after we already have it.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
We iterated over all bgp_path_info's, but once we remove the path, we didn't
check for other paths under the same bgp_dest.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
Currently the code to check to see if any input filters are
applicable is *before* the RFC 8212 check to see if we have
any filters at all. As such we have already tested for this,
so let's move this check for RFC 8212 to immediately before
the input filter test.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
The nexthop reachability code was cut-n-pasted 2 times
with just a tiny bit of difference. If we ever change
that it becomes `fun` to keep them in sync. Since this
is more important than full on speed of code let's abstract
and get bgp_update() to be a bit easier to maintain.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
The code is just arbitrarily checking to see if there are any
mac addresses associated with a prefix. This makes no
sense from the perspective that it can only happen as
an evpn route. Let's not make non-evpn people pay
the price to check this data.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
The variable is_loop_check is being set and then later
we test against it multiple times. Move the setting
of whether or not to check for as loops to where it
is tested against and stop testing it multiple times.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
In the interest of speeding up code, there is no point in
attempting to see if a label is usable if the number of labels
passed in is 0. Since that is a much much quicker test than
the bgp_is_valid_label() call, let's test that first.
Additionally, there is no point in walking the label[] array
passed in unless we are in the if statement, so move it inside.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
In bgp_update(), the two variables allowas_in and aspath_loop_count
are only used when peer->change_local_as is true. Move the retrieval
of the allowas_in data to inside the if statement to save some
(very) small amount of time in bgp_update not gathering this
data unless the particular peer has this set.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
When you have multiple paths to a particular route and a single
path changes. In addition of the other paths are either in
hold down or not established or really just not selected you
could end up with a situation where the bestpath choosen
was a path that was in hold down.
Modify the code such that when there is nothing worse
in bestpath selection for the choosen path, but were
unable to do any sorting, just put the path on the top
of the list and declare it the winner. Else just
do the original and put it at the end.
Signed-off-by: Chirag Shah <chirag@nvidia.com>
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
Fix the display of the local label in show bgp.
> r1# show bgp ipv4 labeled-unicast 172.16.2.2/32
> BGP routing table entry for 172.16.2.2/32, version 2
> Local label: 16 <---- MISSING
> Paths: (1 available, best #1, table default, vrf (null))
> Advertised to non peer-group peers:
> 192.168.1.2
> 65501
> 192.168.1.2 from 192.168.1.2 (172.16.2.2)
> Origin IGP, metric 0, valid, external, best (First path received)
> Remote label: 3
> Last update: Fri Oct 25 17:55:45 2024
Fixes: 67f67ba481 ("bgpd: Drop label_ntop/label_pton functions")
Signed-off-by: Louis Scalbert <louis.scalbert@6wind.com>
This is just a small optimization but when calling path_info_cmp
hundreds of millions of times this adds up.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
bgp_update is a very expensive call. Calling evpn_overlay_free
even when we have no evpn data to free is not trivial. Let's
limit the call into this function until we actually have data to
free.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
For consistency between RIB and BGP, the aigp comparison should
be made after the local route check in bgp bestpath selection.
Signed-off-by: Enke Chen <enchen@paloaltonetworks.com>
Extended communities can be transitive or non-transitive.
Like other attributes (e.g., MED) non-transitive extended communities SHOULD
be sent to the direct peer, but not forward them to eBGP peers next.
Before this patch, we never send non-transitive extended attributes to the
direct peers at all.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
Currently the AIGP is always incremented when a route with the
attribute is advertised. That is incorrect when the nexthop is
unchanged, as is commonly the case in route reflection.
Adjust the AIGP for propagation only when the nexthop is set
to ourselves.
Signed-off-by: Enke Chen <enchen@paloaltonetworks.com>
OAD is treated as an _internal_ BGP peer, and some of the rules (including BGP
attributes) can be relaxed.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
Fix several issues in sourcing AIGP attribute:
1) AIGP should not be set as default for a redistributed route or a
static network. It should be set by config instead.
2) AIGP sourced by "set aigp-metric igp-metric" in a route-map does
not set the correct value for a redistributed route.
3) When redistribute a connected route like loopback, the AGIP (with
value 0) is sourced by "set aigp-metric igp-metric", but the
attribute is not propagated as the attribute flag is not set.
Signed-off-by: Enke Chen <enchen@paloaltonetworks.com>
If the underlay IGP metric changes, we SHOULD re-announce the routes with the
correct bpi->extra->igpmetric set.
Without this patch if the IGP link cost (metric) changes, we never notice this
and the peers do not have the updated metrics, which in turn causes incorrect
best path selections on remote peers.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
The nexthop metric should be added to AIGP when calculating the
bestpath in bgp_path_info_cmp().
Signed-off-by: Enke Chen <enchen@paloaltonetworks.com>
AS 65000 | AS 65001
|
RR |
| |
R1 --- | --- R2
|
When r1 peer is an iBGP route reflector client of rr and r2 peer is a
eBGP neighbor of rr, and all three routers shares the same network, r2
receives announcements coming from r1 with a IPv6 link-local nexthop
from rr. This is incorrect as r2 should send traffic to r1 without
involving rr.
Do not send an IPv6 link-local nexthop if the originating peer is a
route-reflector client.
Link: https://github.com/FRRouting/frr/pull/16219#issuecomment-2397425505
Link: https://github.com/FRRouting/frr/pull/17037#discussion_r1792529683
Signed-off-by: Louis Scalbert <louis.scalbert@6wind.com>
In subgroup_announce_check(), the variable reflect is misleading, as it
suggests a relation to route reflection. However, it actually refers to
the scenario where an iBGP peer announces a route to another iBGP peer.
Rename reflect to ibgp_to_ibgp.
Signed-off-by: Louis Scalbert <louis.scalbert@6wind.com>
If the "nexthop-local unchanged" setting is enabled, it preserves the
IPv6 link-local nexthop from the originating peer. However, if the
originating and destination peers are not on the same network segment,
the originating peer's IPv6 link-local address will be unreachable from
the destination peer.
In such cases, reset the IPv6 link-local nexthop, even if "nexthop-local
unchanged" is set on the destination peer.
Signed-off-by: Louis Scalbert <louis.scalbert@6wind.com>
Do not add an IPv6 link-local nexthop if the originating peer does not
provide one and the nexthop-local unchanged setting is enabled.
Signed-off-by: Louis Scalbert <louis.scalbert@6wind.com>
Currently bgp multipath has these properties:
a) mp_info may or may not be on a single path, based
upon path perturbations in the past.
b) mp_info->count started counting at 0( meaning 1 ). As that the
bestpath path_info was never included in the count
c) The first mp_info in the list held the multipath data associated
with the multipath. As such if you were at any other node that data
was not filled in.
d) As such the mp_info's that are not first on the list basically
were just pointers to the corresponding bgp_path_info that was in
the multipath.
e) On bestpath calculation, a linklist(struct linklist *) of bgp_path_info's was
created.
f) This linklist was passed in to a comparison function that took the
old mpinfo list and compared it item by item to the linklist and
doing magic to figure out how to create a new mp_info list.
g) the old mp_info and the link list had to be memory managed and
freed up.
h) BGP_PATH_MULTIPATH is only set on non bestpath nodes in the
multipath.
This is really complicated. Let's change the algorithm to this:
a) When running bestpath, mark a bgp_path_info node that could be in the ecmp path as
BGP_PATH_MULTIPATH_NEW.
b) When running multipath, just walk the list of bgp_path_info's and if
it has BGP_PATH_MULTIPATH_NEW on it, decide if it is in BGP_MULTIPATH.
If we run out of space to put in the ecmp, clear the flag on the rest.
c) Clean up the counting of sometimes adding 1 to the mpath count.
d) Only allocate a mpath_info node for the bestpath. Clean it up
when done with it.
e) remove the unneeded list management associated with the linklist and
the mp_list.
This greatly simplifies multipath computation for bgp and reduces memory
load for large scale deployments.
2 full feeds in work_queue_run prior:
0 56367.471 1123 50193 493695 50362 493791 0 0 0 TE work_queue_run
BGP multipath info : 1941844 48 110780992 1941844 110780992
2 full feeds in work_queue_run after change:
1 52924.931 1296 40837 465968 41025 487390 0 0 1 TE work_queue_run
BGP multipath info : 970860 32 38836880 970866 38837120
Aproximately 4 seconds of saved cpu time for convergence and ~75 mb
smaller run time.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
This is handy when you need to do source matching e.g. `match src-peer ...`
on outgoing direction with a route-map.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
