May 2022 - Bird-users - Mailing Lists

Scaling BFD support
by Douglas Fischer 24 Jun '22

24 Jun '22

The question raised by colleague Irene reminded me of a topic that may or may not be the focus of BIRD's development. I imagine that the biggest supporters of SMP/Multi-Core/Thread-Safe evolution on BIRD are Operators of Route-Servers of large IXPs, and operators of large-scale Route-Reflectors. Although BFD has its greatest use in the transport network and Underlay, it is increasingly common to see the use of BFD in BGP Internet. I'm personally overly excited about what BIRD version 3 is demonstrating in terms of vertical scalability. But I keep imagining that, even having scalability in the BGP engine, it is almost prohibitive to use BFD in a scenario with a thousand BGP Peers. Is there any view from the IBRD development team for this matter? Or even... Is there any open project focused on BFD that can address this? And going a little further... In a quick Brainstorming... What would be the techniques that could help a Route-Server or Route-Reflector with a thousand BGP Peers to also support a thousand BFD sessions effectively and efficiently? Perhaps resort to some hardware-based off-loading method? Would eBPF help this in any way? -- Douglas Fernando Fischer Engº de Controle e Automação

6 8

Using Babel.router_id in filters
by James Oakley 30 May '22

30 May '22

Hi, I would like to write a filter for a Babel protocol where I set a value based on the Babel router ID. I can see Babel.router_id in the route attributes, but it does not look like this is exposed to filters. Only babel_metric is exposed. Is there a reason for this or was it an oversight? -- James Oakley james(a)ttgi.io

3 2

Way to store ROA info so we can accept but view?
by Dan Mahoney (Gushi) 30 May '22

30 May '22

Hey all, We're using RPKI in testing at the day job, and for a given route, it seems the best we can do is apply a community to it so we can see that it's invalid. A howto I've found says that this is a bad idea and can cause problems. (https://bgpfilterguide.nlnog.net/guides/reject_invalids/) When you look at routes with something like "show route all", there's no field for the RPKI status or the ASes for which ROAs are allowed. So, the questions here is: 1) My understanding of the way RPKI-RTR works is that it's basically handed a tuple of prefix and AS, and RTR says "valid", "invalid", or "unknown". It feels like to check for AS 0 ROAs, we'd basically have to do two lookups for each route that's otherwise invalid, which feels inefficient. Is there a better way? 2) Can the output of "show route" be extended to include user defined fields, or are we locked into what's there? 3) If not, we're limited to adding communities or MEDs or local prefs or something like that, which is a hack, but at least gives us some info we can view. Is that a dangerous trade off? -Dan -- --------Dan Mahoney-------- Techie, Sysadmin, WebGeek Gushi on efnet/undernet IRC FB: fb.com/DanielMahoneyIV LI: linkedin.com/in/gushi Site: http://www.gushi.org ---------------------------

6 12

scalability and <RMT> error
by dytadam＠sina.com 29 May '22

29 May '22

Hi there, I was running OSPFv2 for a scale of more than a thousand. However, I noticed that the routing table keeps changing after running for hours. It seems the routing protocol has not come to stable point. And the log says <RMT> Bad DBDES packet from nbr, too late for DD exchange. How can I solve it? Is it because the packet is too large? Thank you so much.

2 1

[PATCH] RPKI: Display cache server port on show protocol
by Luiz Amaral 27 May '22

27 May '22

Hello, Recently I needed to check the address and port for my cache server, only to find out that the "show protocol all rpki1" output omitted the port. This small patch appends the port after the cache server address on the show protocol all output. Thanks, Luiz

2 1

[PATCH] RPKI: Implement VRF support
by Luiz Amaral 27 May '22

27 May '22

Hello, I made a small patch that adds VRF support to the RPKI protocol. I did some basic testing with and without VRF and it worked as expected. Please let me know if I missed anything. Thanks, Luiz

2 1

[PATCH] babel: Don't try to remove multicast seqno request objects from neighbour list
by Toke Høiland-Jørgensen 27 May '22

27 May '22

The Babel seqno request code keeps track of which seqno requests are outstanding for a neighbour by putting them onto a per-neighbour list. When reusing a seqno request, it will try to remove this node, but if the seqno request in question was a multicast request with no neighbour attached this will result in a crash because it tries to remove a list node that wasn't added to any list. Fix this by making the list remove conditional. Also add a check so that seqno requests are only reused if the neighbour also matches, allowing multiple outstanding requests for the same router ID. Fixes: ebd5751cdeb4 ("Babel: Seqno requests are properly decoupled from neighbors when the underlying interface disappears") Reported-by: Stefan Haller <stefan.haller(a)stha.de> Signed-off-by: Toke Høiland-Jørgensen <toke(a)toke.dk> --- proto/babel/babel.c | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) diff --git a/proto/babel/babel.c b/proto/babel/babel.c index 4a7d550f5efe..03f1cb500278 100644 --- a/proto/babel/babel.c +++ b/proto/babel/babel.c @@ -304,7 +304,7 @@ babel_add_seqno_request(struct babel_proto *p, struct babel_entry *e, struct babel_seqno_request *sr; WALK_LIST(sr, e->requests) - if (sr->router_id == router_id) + if (sr->router_id == router_id && sr->nbr == nbr) { /* Found matching or newer */ if (ge_mod64k(sr->seqno, seqno) && seqno_request_valid(sr)) @@ -312,7 +312,8 @@ babel_add_seqno_request(struct babel_proto *p, struct babel_entry *e, /* Found older */ rem_node(NODE sr); - rem_node(&sr->nbr_node); + if (sr->nbr) + rem_node(&sr->nbr_node); goto found; } @@ -349,17 +350,20 @@ static int babel_satisfy_seqno_request(struct babel_proto *p, struct babel_entry *e, u64 router_id, u16 seqno) { - struct babel_seqno_request *sr; + struct babel_seqno_request *sr, *srx; + int ret = 0; - WALK_LIST(sr, e->requests) + WALK_LIST_DELSAFE(sr, srx, e->requests) if ((sr->router_id == router_id) && ge_mod64k(seqno, sr->seqno)) { - /* Found the request, remove it */ + /* Found a matching request, remove it; there may be multiple outstanding + * requests, so continue looping + */ babel_remove_seqno_request(p, sr); - return 1; + ret = 1; } - return 0; + return ret; } static void -- 2.36.1

2 1

[PATCH 1/2] Babel: Add option to support ecmp routes
by Daniel Gröber 26 May '22

26 May '22

--- The FIB walks in babel_reconfigure_iface are a bit ugly, I tried to loop over ifa->neigh_list instead but couldn't get it to work. I'm open to suggestions :) doc/bird.sgml | 8 +++++ proto/babel/babel.c | 70 +++++++++++++++++++++++++++++++++++++------- proto/babel/babel.h | 3 ++ proto/babel/config.Y | 3 ++ 4 files changed, 73 insertions(+), 11 deletions(-) diff --git a/doc/bird.sgml b/doc/bird.sgml index 1580facd..5e85d8ec 100644 --- a/doc/bird.sgml +++ b/doc/bird.sgml @@ -1879,6 +1879,7 @@ protocol babel [<name>] { check link <switch>; next hop ipv4 <address>; next hop ipv6 <address>; + ecmp <switch> [limit <num>]; authentication none|mac [permissive]; password "<text>"; password "<text>" { @@ -1983,6 +1984,13 @@ protocol babel [<name>] { source for Babel packets will be used. In normal operation, it should not be necessary to set this option. + <tag><label id="babel-ecmp">ecmp <m>switch</m> [limit <m>number</m>]</tag> + + Determines whether babel will emit ECMP (equal-cost multipath) + routes, allowing to load-balancing traffic across multiple paths. If + enabled the maximum number of next-hops to allow can be specified, + defaulting to 16. + <tag><label id="babel-authentication">authentication none|mac [permissive]</tag> Selects authentication method to be used. <cf/none/ means that packets are not authenticated at all, <cf/mac/ means MAC authentication is diff --git a/proto/babel/babel.c b/proto/babel/babel.c index 4a7d550f..4d9c657c 100644 --- a/proto/babel/babel.c +++ b/proto/babel/babel.c @@ -623,6 +623,34 @@ done: } } +/** + * babel_nexthop_insert - add next_hop of route to nexthop list + * @p: Babel protocol instance + * @r: Babel route + * @nhs: nexthop list head to append onto + * @nh: freshly allocated buffer to fill + */ +static void +babel_nexthop_insert( + struct babel_proto *p, + struct babel_route *r, + struct nexthop **nhs, + struct nexthop *nh) +{ + nh->gw = r->next_hop; + nh->iface = r->neigh->ifa->iface; + + /* + * If we cannot find a reachable neighbour, set the entry to be onlink. This + * makes it possible to, e.g., assign /32 addresses on a mesh interface and + * have routing work. + */ + if (!neigh_find(&p->p, r->next_hop, r->neigh->ifa->iface, 0)) + nh->flags = RNF_ONLINK; + + nexthop_insert(nhs, nh); +} + /** * babel_announce_rte - announce selected route to the core * @p: Babel protocol instance @@ -635,6 +663,7 @@ done: static void babel_announce_rte(struct babel_proto *p, struct babel_entry *e) { + struct babel_config *cf = (void *) p->p.cf; struct babel_route *r = e->selected; struct channel *c = (e->n.addr->type == NET_IP4) ? p->ip4_channel : p->ip6_channel; @@ -645,18 +674,24 @@ babel_announce_rte(struct babel_proto *p, struct babel_entry *e) .source = RTS_BABEL, .scope = SCOPE_UNIVERSE, .dest = RTD_UNICAST, - .from = r->neigh->addr, - .nh.gw = r->next_hop, - .nh.iface = r->neigh->ifa->iface, }; - /* - * If we cannot find a reachable neighbour, set the entry to be onlink. This - * makes it possible to, e.g., assign /32 addresses on a mesh interface and - * have routing work. - */ - if (!neigh_find(&p->p, r->next_hop, r->neigh->ifa->iface, 0)) - a0.nh.flags = RNF_ONLINK; + struct nexthop *nhs = NULL; + babel_nexthop_insert(p, r, &nhs, allocz(sizeof(struct nexthop))); + int num_nexthops = 1; + + struct babel_route *cr; + WALK_LIST(cr, e->routes) { + if (num_nexthops++ >= cf->ecmp) + break; + + if (cr == r || !cr->feasible || cr->metric != r->metric) + continue; + + babel_nexthop_insert(p, cr, &nhs, allocz(sizeof(struct nexthop))); + } + + a0.nh = *nhs; rta *a = rta_lookup(&a0); rte *rte = rte_get_temp(a); @@ -736,6 +771,7 @@ babel_announce_retraction(struct babel_proto *p, struct babel_entry *e) static void babel_select_route(struct babel_proto *p, struct babel_entry *e, struct babel_route *mod) { + struct babel_config *cf = (void *) p->p.cf; struct babel_route *r, *best = e->selected; /* Shortcut if only non-best was modified */ @@ -744,8 +780,10 @@ babel_select_route(struct babel_proto *p, struct babel_entry *e, struct babel_ro /* Either select modified route, or keep old best route */ if ((mod->metric < (best ? best->metric : BABEL_INFINITY)) && mod->feasible) best = mod; - else + else if (cf->ecmp==1) return; + /* With ecmp one of the non-selected but equal metric routes might have + * changed so contine on with the announcement in that case. */ } else { @@ -1879,6 +1917,16 @@ babel_reconfigure_iface(struct babel_proto *p, struct babel_iface *ifa, struct b if (ifa->up) babel_iface_kick_timer(ifa); + /* Update all routes to refresh ecmp settings. */ + FIB_WALK(&p->ip6_rtable, struct babel_entry, e) { + babel_select_route(p, e, NULL); + } + FIB_WALK_END; + FIB_WALK(&p->ip4_rtable, struct babel_entry, e) { + babel_select_route(p, e, NULL); + } + FIB_WALK_END; + return 1; } diff --git a/proto/babel/babel.h b/proto/babel/babel.h index 84feb085..3868dcb2 100644 --- a/proto/babel/babel.h +++ b/proto/babel/babel.h @@ -62,6 +62,8 @@ #define BABEL_OVERHEAD (IP6_HEADER_LENGTH+UDP_HEADER_LENGTH) #define BABEL_MIN_MTU (512 + BABEL_OVERHEAD) +#define BABEL_DEFAULT_ECMP_LIMIT 16 + #define BABEL_AUTH_NONE 0 #define BABEL_AUTH_MAC 1 @@ -120,6 +122,7 @@ struct babel_config { list iface_list; /* List of iface configs (struct babel_iface_config) */ uint hold_time; /* Time to hold stale entries and unreachable routes */ u8 randomize_router_id; + u8 ecmp; /* Maximum number of nexthops to install */ struct channel_config *ip4_channel; struct channel_config *ip6_channel; diff --git a/proto/babel/config.Y b/proto/babel/config.Y index 05210fa4..8f18c790 100644 --- a/proto/babel/config.Y +++ b/proto/babel/config.Y @@ -36,6 +36,7 @@ babel_proto_start: proto_start BABEL this_proto = proto_config_new(&proto_babel, $1); init_list(&BABEL_CFG->iface_list); BABEL_CFG->hold_time = 1 S_; + BABEL_CFG->ecmp = 1; }; babel_proto_item: @@ -43,6 +44,8 @@ babel_proto_item: | proto_channel | INTERFACE babel_iface | RANDOMIZE ROUTER ID bool { BABEL_CFG->randomize_router_id = $4; } + | ECMP bool { BABEL_CFG->ecmp = $2 ? BABEL_DEFAULT_ECMP_LIMIT : 1; } + | ECMP bool LIMIT expr { BABEL_CFG->ecmp = $2 ? $4 : 1; } ; babel_proto_opts: -- 2.30.2

3 10

[PATCH] Netlink: Propagate ecmp nexthop weight to kernel for inet6 routes
by Daniel Gröber 22 May '22

22 May '22

Previously nl_send_route would use plain nl_add_nexthop for ecmp ipv6 routes instead of adding RTA_MULTIPATH objects via nl_add_multipath. The former lacks support for the rtnexthop.rtnh_hops field need for setting the nexthop weight though. On the kernel side support for nexthop weights was introduced by commit 398958ae48 ("ipv6: Add support for non-equal-cost multipath") which landed in 4.16. Before this the weight will simply be ignored. Support for the ipv6 RTA_MULTIPATH attribute was added in 51ebd31815 ("ipv6: add support of equal cost multipath (ECMP)") which landed in 3.10. If this version bound isn't met installing the route will fail. --- sysdep/linux/netlink.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sysdep/linux/netlink.c b/sysdep/linux/netlink.c index 29b744cb..5e0b036d 100644 --- a/sysdep/linux/netlink.c +++ b/sysdep/linux/netlink.c @@ -1446,7 +1446,7 @@ dest: { case RTD_UNICAST: r->r.rtm_type = RTN_UNICAST; - if (nh->next && !krt_ecmp6(p)) + if (nh->next) nl_add_multipath(&r->h, rsize, nh, p->af, eattrs); else { -- 2.30.2

1 0

Bird2 and protocol Babel. auto change default gw
by Michele Salerno 16 May '22

16 May '22

Hi, I am new to ML and also to using bird2. I use bird2 with babel protocol for routing the vpn server and openwrt router. The openwrt routers have a mesh wifi interface. My question is, if the wan (default gw) is down, can bird use the better mesh as the default gw? If the wan is restored, can it come back as default gw? Best regards, Michele

2 1