diff options
| author | Philippe Guibert <philippe.guibert@6wind.com> | 2024-04-09 17:09:39 +0200 |
|---|---|---|
| committer | Philippe Guibert <philippe.guibert@6wind.com> | 2024-04-18 14:55:25 +0200 |
| commit | 8b27521cfa0c93dc326698d9f729a0486fc43ead (patch) | |
| tree | 7c242ccf96c89bc114a8184226ba2f408f8942aa /bgpd/bgp_mplsvpn.c | |
| parent | Merge pull request #15770 from idryzhov/fix-yang-exit (diff) | |
| download | frr-8b27521cfa0c93dc326698d9f729a0486fc43ead.tar.xz frr-8b27521cfa0c93dc326698d9f729a0486fc43ead.zip | |
bgpd: do allocate vrf label only when necessary
Today, with the following bgp instance configured, the
local VRF label is allocated even if it is not used.
> router bgp 65500 vrf vrf1
> address-family ipv4 unicast
> label vpn export allocation-mode per-nexthop
> label vpn export auto
> rd vpn export 444:1
> rt vpn both 52:100
> export vpn
> import vpn
The 'show mpls table' indicates that the 16 label value
is allocated, but never used in the exported prefixes.
> r1# show mpls table
> Inbound Label Type Nexthop Outbound Label
> -----------------------------------------------------
> 16 BGP vrf1 -
> 17 BGP 192.168.255.13 -
> 18 BGP 192.0.2.12 -
> 19 BGP 192.0.2.11 -
Fix this by only allocating new label values when really
used. Consequently, only 3 labels will be allocated instead
of previously 4.
> r1# show mpls table
> Inbound Label Type Nexthop Outbound Label
> -----------------------------------------------------
> 16 BGP 192.168.255.13 -
> 17 BGP 192.0.2.11 -
> 18 BGP 192.0.2.12 -
Fixes: 577be36a41be ("bgpd: add support for l3vpn per-nexthop label")
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
Diffstat (limited to 'bgpd/bgp_mplsvpn.c')
| -rw-r--r-- | bgpd/bgp_mplsvpn.c | 28 |
1 files changed, 18 insertions, 10 deletions
diff --git a/bgpd/bgp_mplsvpn.c b/bgpd/bgp_mplsvpn.c index 2404ceffb..d17fdc716 100644 --- a/bgpd/bgp_mplsvpn.c +++ b/bgpd/bgp_mplsvpn.c @@ -1441,6 +1441,16 @@ _vpn_leak_from_vrf_get_per_nexthop_label(struct bgp_path_info *pi, return blnc->label; } +static mpls_label_t bgp_mplsvpn_get_vpn_label(struct vpn_policy *bgp_policy) +{ + if (bgp_policy->tovpn_label == MPLS_LABEL_NONE && + CHECK_FLAG(bgp_policy->flags, BGP_VPN_POLICY_TOVPN_LABEL_AUTO)) { + bgp_lp_get(LP_TYPE_VRF, bgp_policy, vpn_leak_label_callback); + return MPLS_INVALID_LABEL; + } + return bgp_policy->tovpn_label; +} + /* Filter out all the cases where a per nexthop label is not possible: * - return an invalid label when the nexthop is invalid * - return the per VRF label when the per nexthop label is not supported @@ -1469,7 +1479,7 @@ vpn_leak_from_vrf_get_per_nexthop_label(afi_t afi, struct bgp_path_info *pi, * Fallback to the per VRF label. */ bgp_mplsvpn_path_nh_label_unlink(pi); - return from_bgp->vpn_policy[afi].tovpn_label; + return bgp_mplsvpn_get_vpn_label(&from_bgp->vpn_policy[afi]); } if (is_bgp_static_route == false && afi == AFI_IP && @@ -1481,7 +1491,7 @@ vpn_leak_from_vrf_get_per_nexthop_label(afi_t afi, struct bgp_path_info *pi, * Fallback to the per VRF label. */ bgp_mplsvpn_path_nh_label_unlink(pi); - return from_bgp->vpn_policy[afi].tovpn_label; + return bgp_mplsvpn_get_vpn_label(&from_bgp->vpn_policy[afi]); } if (is_bgp_static_route == false && afi == AFI_IP6 && @@ -1495,7 +1505,7 @@ vpn_leak_from_vrf_get_per_nexthop_label(afi_t afi, struct bgp_path_info *pi, * Fallback to the per VRF label. */ bgp_mplsvpn_path_nh_label_unlink(pi); - return from_bgp->vpn_policy[afi].tovpn_label; + return bgp_mplsvpn_get_vpn_label(&from_bgp->vpn_policy[afi]); } /* Check the next-hop reachability. @@ -1517,7 +1527,7 @@ vpn_leak_from_vrf_get_per_nexthop_label(afi_t afi, struct bgp_path_info *pi, * table. Fallback to the per-vrf label */ bgp_mplsvpn_path_nh_label_unlink(pi); - return from_bgp->vpn_policy[afi].tovpn_label; + return bgp_mplsvpn_get_vpn_label(&from_bgp->vpn_policy[afi]); } if (!nh_valid || !pi->nexthop || pi->nexthop->nexthop_num == 0 || @@ -1540,7 +1550,7 @@ vpn_leak_from_vrf_get_per_nexthop_label(afi_t afi, struct bgp_path_info *pi, * Fallback to per-vrf label. */ bgp_mplsvpn_path_nh_label_unlink(pi); - return from_bgp->vpn_policy[afi].tovpn_label; + return bgp_mplsvpn_get_vpn_label(&from_bgp->vpn_policy[afi]); } return _vpn_leak_from_vrf_get_per_nexthop_label(pi, to_bgp, from_bgp, @@ -1755,12 +1765,10 @@ void vpn_leak_from_vrf_update(struct bgp *to_bgp, /* to */ label_val = vpn_leak_from_vrf_get_per_nexthop_label( afi, path_vrf, from_bgp, to_bgp); else - /* per VRF label mode */ - label_val = from_bgp->vpn_policy[afi].tovpn_label; + label_val = + bgp_mplsvpn_get_vpn_label(&from_bgp->vpn_policy[afi]); - if (label_val == MPLS_INVALID_LABEL && - CHECK_FLAG(from_bgp->vpn_policy[afi].flags, - BGP_VPN_POLICY_TOVPN_LABEL_PER_NEXTHOP)) { + if (label_val == MPLS_INVALID_LABEL) { /* no valid label for the moment * when the 'bgp_mplsvpn_get_label_per_nexthop_cb' callback gets * a valid label value, it will call the current function again. |