November 2011 - Bird-users

[PATCH 1/1] * Implement correct RFC4271 MED comparison
by Alexander V. Chernikov 18 Nov '11

18 Nov '11

--- filter/filter.c | 6 + nest/proto-hooks.c | 3 +- nest/protocol.h | 6 +- nest/route.h | 8 ++ nest/rt-table.c | 58 +++++++++-- proto/bgp/attrs.c | 289 +++++++++++++++++++++++++++++++++++++++++++++++++++- proto/bgp/bgp.c | 4 +- proto/bgp/bgp.h | 9 ++- proto/ospf/ospf.c | 4 +- proto/rip/rip.c | 4 +- 10 files changed, 374 insertions(+), 17 deletions(-) diff --git a/filter/filter.c b/filter/filter.c index b3b17dc..77af39a 100644 --- a/filter/filter.c +++ b/filter/filter.c @@ -935,6 +935,12 @@ interpret(struct f_inst *what) if (v1.type != T_PATH) runtime( "Setting path attribute to non-path value" ); l->attrs[0].u.ptr = v1.val.ad; + /* + * Zero cache on AS-PATH change. + * XXX: This breaks COW concept but + * removing this flag is harmless + */ + (*f_rte)->pflags &= ~0x01; break; case EAF_TYPE_INT_SET: if (v1.type != T_CLIST) diff --git a/nest/proto-hooks.c b/nest/proto-hooks.c index f026192..1dec95c 100644 --- a/nest/proto-hooks.c +++ b/nest/proto-hooks.c @@ -271,6 +271,7 @@ int import_control(struct proto *p, rte **e, ea_list **attrs, struct linpool *po * rte_better - compare metrics of two routes * @new: the new route * @old: the original route + * @flags: protocol-specific comparison flags * * This hook gets called when the routing table contains two routes * for the same network which have originated from different instances @@ -279,7 +280,7 @@ int import_control(struct proto *p, rte **e, ea_list **attrs, struct linpool *po * * Result: 1 if @new is better (more preferred) than @old, 0 otherwise. */ -int rte_better(rte *new, rte *old) +int rte_better(rte *new, rte *old, u32 flags) { DUMMY; } /** diff --git a/nest/protocol.h b/nest/protocol.h index a7518c2..02defb4 100644 --- a/nest/protocol.h +++ b/nest/protocol.h @@ -178,13 +178,15 @@ struct proto { /* * Routing entry hooks (called only for rte's belonging to this protocol): * - * rte_better Compare two rte's and decide which one is better (1=first, 0=second). + * rte_best Find best rte starting from given. + * rte_better Compare two rte's and decide if better (1=first, 0=second). * rte_same Compare two rte's and decide whether they are identical (1=yes, 0=no). * rte_insert Called whenever a rte is inserted to a routing table. * rte_remove Called whenever a rte is removed from the routing table. */ - int (*rte_better)(struct rte *, struct rte *); + struct rte *(*rte_best)(struct rte *); + int (*rte_better)(struct rte *, struct rte *, u32 flags); int (*rte_same)(struct rte *, struct rte *); void (*rte_insert)(struct network *, struct rte *); void (*rte_remove)(struct network *, struct rte *); diff --git a/nest/route.h b/nest/route.h index a4c0154..31a7a56 100644 --- a/nest/route.h +++ b/nest/route.h @@ -201,6 +201,12 @@ typedef struct rte { u32 router_id; /* Router that originated this route */ } ospf; #endif +#ifdef CONFIG_BGP + struct { + u32 first_as; /* First as number in AS-PATH or local AS */ + struct rte *next_item; /* Used in bgp_get_best */ + } bgp; +#endif struct { /* Routes generated by krt sync (both temporary and inherited ones) */ s8 src; /* Alleged route source (see krt.h) */ u8 proto; /* Kernel source protocol ID */ @@ -212,6 +218,8 @@ typedef struct rte { } rte; #define REF_COW 1 /* Copy this rte on write */ +/* Protocol-specific flags */ +#define PFR_FLAG1 0x01 /* Definition of first possible flag */ /* Types of route announcement, also used as flags */ #define RA_OPTIMAL 1 /* Announcement of optimal route change */ diff --git a/nest/rt-table.c b/nest/rt-table.c index e20d2f6..574711a 100644 --- a/nest/rt-table.c +++ b/nest/rt-table.c @@ -135,10 +135,23 @@ rte_do_cow(rte *r) return e; } -static int /* Actually better or at least as good as */ +/** + * rte_better - returns 1 if new rte is better + * @new: new rte + * @old: old BEST rte + * + * Functions compares old and new rte returning 1 IFF new rte is better. + * Best rte in @old->net is required to be specified in @old if @old and @new are + * originated from same protocol type. Specifying different rte can produce + * incorrect calculations in some protocols (see bgp_rte_better) + * + * Returns 1 if @new is better, 0 overwise + */ +static int rte_better(rte *new, rte *old) { - int (*better)(rte *, rte *); + int (*better)(rte *, rte *, u32); + struct proto *p_old, *p_new; if (!old) return 1; @@ -146,17 +159,19 @@ rte_better(rte *new, rte *old) return 1; if (new->pref < old->pref) return 0; - if (new->attrs->proto->proto != old->attrs->proto->proto) + p_old = old->attrs->proto; + p_new = new->attrs->proto; + if (p_new->proto != p_old->proto) { /* * If the user has configured protocol preferences, so that two different protocols * have the same preference, try to break the tie by comparing addresses. Not too * useful, but keeps the ordering of routes unambiguous. */ - return new->attrs->proto->proto > old->attrs->proto->proto; + return p_new->proto > p_old->proto; } - if (better = new->attrs->proto->rte_better) - return better(new, old); + if (better = p_new->rte_better) + return better(new, old, 0); return 0; } @@ -426,9 +441,11 @@ static void rte_recalculate(rtable *table, net *net, struct proto *p, struct proto *src, rte *new, ea_list *tmpa) { struct proto_stats *stats = &p->stats; + struct proto *rte_proto; rte *old_best = net->routes; rte *old = NULL; - rte **k, *r, *s; + rte **k, *r, *s, *t; + u32 proto_mask, attr_class; #ifdef CONFIG_PIPE if (proto_is_pipe(p) && (p->table == table)) @@ -526,9 +543,36 @@ rte_recalculate(rtable *table, net *net, struct proto *p, struct proto *src, rte /* Find new optimal route */ r = NULL; + proto_mask = 0; for (s=net->routes; s; s=s->next) + { + rte_proto = s->attrs->proto; + attr_class = rte_proto->proto->attr_class; + /* Protocol has non-zero attr_class and rte_best hook */ + if (attr_class && rte_proto->rte_best) + { + /* Check if this is first hook invocation */ + if (!((2 << attr_class) & proto_mask)) + { + proto_mask |= 2 << attr_class; + /* + * Find best route from protocol point of view + * and compare it with current best route if exists + */ + t = rte_proto->rte_best(s); + if (!r || rte_better(r, t)) + r = t; + } + /* + * Protocol has already supplied its best route. + * Skip this rte + */ + continue; + } + if (rte_better(s, r)) r = s; + } /* Announce optimal route */ rte_announce(table, RA_OPTIMAL, net, r, old_best, tmpa); diff --git a/proto/bgp/attrs.c b/proto/bgp/attrs.c index 2832f42..0b417df 100644 --- a/proto/bgp/attrs.c +++ b/proto/bgp/attrs.c @@ -1118,7 +1118,7 @@ rte_resolvable(rte *rt) } int -bgp_rte_better(rte *new, rte *old) +bgp_rte_better(rte *new, rte *old, u32 flags UNUSED) { struct bgp_proto *new_bgp = (struct bgp_proto *) new->attrs->proto; struct bgp_proto *old_bgp = (struct bgp_proto *) old->attrs->proto; @@ -1236,6 +1236,293 @@ bgp_rte_better(rte *new, rte *old) return (ipa_compare(new_bgp->cf->remote_ip, old_bgp->cf->remote_ip) < 0); } + + +static inline u32 +bgp_get_neighbor_as(rte *s) +{ + u32 first_as; + + if (s->flags & PFR_BGP_ASCACHE) + return s->u.bgp.first_as; + + first_as = bgp_get_neighbor(s); + s->u.bgp.first_as = first_as; + s->pflags |= PFR_BGP_ASCACHE; + + return first_as; +} + +/** + * bgp_rte_better_med - Determine best route + * @new: new rte + * @old: old BEST rte in @old->net + * + * Function implements RFC 4271 best route selection + * with so-called 'deterministic' MED comparison. + * If @new and @old are announced from different neighbours + * @old is required to be the best route. + * + * Returns: 1 if @new is better, 0 overwise + */ +int +bgp_rte_better_med(rte *new, rte *old, u32 flags) +{ + struct bgp_proto *new_bgp = (struct bgp_proto *) new->attrs->proto; + struct bgp_proto *old_bgp = (struct bgp_proto *) old->attrs->proto; + struct bgp_config *new_config, *old_config; + ea_list *new_eattrs = new->attrs->eattrs; + ea_list *old_eattrs = old->attrs->eattrs; + eattr *x, *y; + u32 n, o; + u32 new_as, old_as, tmp_as; + rte *s, *r; + + /* RFC 4271 9.1.2.1. Route resolvability test */ + n = rte_resolvable(new); + o = rte_resolvable(old); + if (n > o) + return 1; + if (n < o) + return 0; + + /* Start with local preferences */ + x = ea_find(new_eattrs, EA_CODE(EAP_BGP, BA_LOCAL_PREF)); + y = ea_find(old_eattrs, EA_CODE(EAP_BGP, BA_LOCAL_PREF)); + n = x ? x->u.data : new_bgp->cf->default_local_pref; + o = y ? y->u.data : old_bgp->cf->default_local_pref; + if (n > o) + return 1; + if (n < o) + return 0; + + /* RFC 4271 9.1.2.2. a) Use AS path lengths */ + if (new_bgp->cf->compare_path_lengths || old_bgp->cf->compare_path_lengths) + { + x = ea_find(new_eattrs, EA_CODE(EAP_BGP, BA_AS_PATH)); + y = ea_find(old_eattrs, EA_CODE(EAP_BGP, BA_AS_PATH)); + n = x ? as_path_getlen(x->u.ptr) : AS_PATH_MAXLEN; + o = y ? as_path_getlen(y->u.ptr) : AS_PATH_MAXLEN; + if (n < o) + return 1; + if (n > o) + return 0; + } + + /* RFC 4271 9.1.2.2. b) Use origins */ + x = ea_find(new_eattrs, EA_CODE(EAP_BGP, BA_ORIGIN)); + y = ea_find(old_eattrs, EA_CODE(EAP_BGP, BA_ORIGIN)); + n = x ? x->u.data : ORIGIN_INCOMPLETE; + o = y ? y->u.data : ORIGIN_INCOMPLETE; + if (n < o) + return 1; + if (n > o) + return 0; + + /* RFC 4271 9.1.2.2. c) Compare MED's */ + /* + * This is a bit tricky. + * + * First we retrieve/update neighbor ASNs from rte cache. + * If @old and @new ASNs are the same, we can immediately + * compare MED. If not, we need to compare best routes + * from both ASNs. Since @old is the best route already nothing + * needs to be done with @old. However, we need to scan all + * routes for given network (another place where @old is required + * to be the best (and the first) route) to get best rte in + * @new neighbor ASN. This changes complexity from O(1) to O(N). + * If found rte is not @new, we simply returns 0 + * since @old is the best rte of all routes. Overwise, we + * continue to compare @old and @new skippind MED comparison. + */ + /* Get neighbor info first as && update cache if nesessary */ + old_as = bgp_get_neighbor_as(old); + new_as = bgp_get_neighbor_as(new); + + new_config = new_bgp->cf; + old_config = old_bgp->cf; + + if (new_as == old_as) + { + /* We can compare MED */ + x = ea_find(new_eattrs, EA_CODE(EAP_BGP, BA_MULTI_EXIT_DISC)); + y = ea_find(old_eattrs, EA_CODE(EAP_BGP, BA_MULTI_EXIT_DISC)); + n = x ? x->u.data : new_config->default_med; + o = y ? y->u.data : old_config->default_med; + if (n < o) + return 1; + if (n > o) + return 0; + } + else + { + /* + * routes came from different neighbor ASses. + * We need to find best route within routes + * from new_as + */ + r = new; + for (s = old; s; s = s->next) + { + /* Get neighbor ASN */ + tmp_as = bgp_get_neighbor_as(s); + if (tmp_as != new_as) + continue; + + /* + * Okay, we've found another rte from the same neighbor. + * Let's check it. + */ + if (bgp_rte_better_med(s, r, 0)) + r = s; + } + + /* + * Best route for neighbor routes is saved in r. + * If new is not the best route in the group this + * means that current best route remains best. + */ + if (r != new) + return 0; + } + + /* RFC 4271 9.1.2.2. d) Prefer external peers */ + if (new_bgp->is_internal > old_bgp->is_internal) + return 0; + if (new_bgp->is_internal < old_bgp->is_internal) + return 1; + + /* RFC 4271 9.1.2.2. e) Compare IGP metrics */ + n = new_config->igp_metric ? new->attrs->igp_metric : 0; + o = old_config->igp_metric ? old->attrs->igp_metric : 0; + if (n < o) + return 1; + if (n > o) + return 0; + + /* RFC 4271 9.1.2.2. f) Compare BGP identifiers */ + /* RFC 4456 9. a) Use ORIGINATOR_ID instead of local neighor ID */ + x = ea_find(new_eattrs, EA_CODE(EAP_BGP, BA_ORIGINATOR_ID)); + y = ea_find(old_eattrs, EA_CODE(EAP_BGP, BA_ORIGINATOR_ID)); + n = x ? x->u.data : new_bgp->remote_id; + o = y ? y->u.data : old_bgp->remote_id; + + /* RFC 5004 - prefer older routes */ + /* (if both are external and from different peer) */ + if ((new_config->prefer_older || old_config->prefer_older) && + !new_bgp->is_internal && n != o) + return 0; + + /* rest of RFC 4271 9.1.2.2. f) */ + if (n < o) + return 1; + if (n > o) + return 0; + + /* RFC 4456 9. b) Compare cluster list lengths */ + x = ea_find(new_eattrs, EA_CODE(EAP_BGP, BA_CLUSTER_LIST)); + y = ea_find(old_eattrs, EA_CODE(EAP_BGP, BA_CLUSTER_LIST)); + n = x ? int_set_get_size(x->u.ptr) : 0; + o = y ? int_set_get_size(y->u.ptr) : 0; + if (n < o) + return 1; + if (n > o) + return 0; + + /* RFC 4271 9.1.2.2. g) Compare peer IP adresses */ + return (ipa_compare(new_config->remote_ip, old_config->remote_ip) < 0); +} + +#define LOCAL_HASH 16 +/** + * bgp_rte_best - returns the best rte in chain originated by BGP + * + */ +rte * +bgp_rte_best(rte *old) +{ + rte **rgroups, *s, *ss, *best = NULL; + u32 local_as, best_as; + int i; + + if (!old->next) + return old; + + /* Allocate hash for ASn groups */ + rgroups = alloca(LOCAL_HASH * sizeof(rte *)); + memset(rgroups, 0, LOCAL_HASH * sizeof(rte *)); + + for (s = old; s; s = s->next) + { + local_as = bgp_get_neighbor_as(s); + best = rgroups[local_as & (LOCAL_HASH - 1)]; + if (!best) + { + /* New group */ + rgroups[local_as & (LOCAL_HASH - 1)] = s; + s->u.bgp.next_item = NULL; + continue; + } + + ss = best; + best_as = bgp_get_neighbor_as(ss); + while (ss->u.bgp.next_item) + { + if ((best_as = bgp_get_neighbor_as(ss)) == local_as) + break; + /* Hash collision */ + ss = ss->u.bgp.next_item; + } + + /* List ended */ + if (best_as == local_as) + { + /* Both rte are from the same group, let's compare */ + if (bgp_rte_better_med(s, best, 0)) + { + /* New best rte in group */ + rgroups[local_as & (LOCAL_HASH - 1)] = s; + s->u.bgp.next_item = best->u.bgp.next_item; + } + + continue; + } + + /* New group */ + ss->u.bgp.next_item = s; + s->u.bgp.next_item = NULL; + } + + /* Cycle ended, now we have to compare all per-group best rtes */ + for (i = 0, best = NULL; i < LOCAL_HASH; i++) + { + for (s = rgroups[i]; s; s = s->u.bgp.next_item) + { + if (!best) + { + best = s; + continue; + } + + /* Do general preference checking */ + if (best->pref > s->pref) + continue; + if (s->pref > best->pref) + { + best = s; + continue; + } + + /* Do full bgp-specific checking without MED comparison */ + if (bgp_rte_better_med(s, best, BGP_CFLAG_SKIPMED)) + best = s; + } + } + + return best; +} +#undef LOCAL_HASH + static struct adata * bgp_aggregator_convert_to_new(struct adata *old, struct linpool *pool) { diff --git a/proto/bgp/bgp.c b/proto/bgp/bgp.c index 675342d..a277680 100644 --- a/proto/bgp/bgp.c +++ b/proto/bgp/bgp.c @@ -904,7 +904,9 @@ bgp_init(struct proto_config *C) P->accept_ra_types = RA_OPTIMAL; P->rt_notify = bgp_rt_notify; - P->rte_better = bgp_rte_better; + //P->rte_better = bgp_rte_better; + P->rte_better = bgp_rte_better_med; + P->rte_best = bgp_rte_best; P->import_control = bgp_import_control; P->neigh_notify = bgp_neigh_notify; P->reload_routes = bgp_reload_routes; diff --git a/proto/bgp/bgp.h b/proto/bgp/bgp.h index 437ba33..70e1ff9 100644 --- a/proto/bgp/bgp.h +++ b/proto/bgp/bgp.h @@ -184,7 +184,9 @@ void bgp_attach_attr(struct ea_list **to, struct linpool *pool, unsigned attr, u byte *bgp_attach_attr_wa(struct ea_list **to, struct linpool *pool, unsigned attr, unsigned len); struct rta *bgp_decode_attrs(struct bgp_conn *conn, byte *a, unsigned int len, struct linpool *pool, int mandatory); int bgp_get_attr(struct eattr *e, byte *buf, int buflen); -int bgp_rte_better(struct rte *, struct rte *); +int bgp_rte_better(struct rte *, struct rte *, u32); +int bgp_rte_better_med(struct rte *, struct rte *, u32); +rte *bgp_rte_best(rte *old); void bgp_rt_notify(struct proto *P, rtable *tbl UNUSED, net *n, rte *new, rte *old UNUSED, ea_list *attrs); int bgp_import_control(struct proto *, struct rte **, struct ea_list **, struct linpool *); void bgp_attr_init(struct bgp_proto *); @@ -241,6 +243,11 @@ void bgp_log_error(struct bgp_proto *p, u8 class, char *msg, unsigned code, unsi #define BA_AS4_PATH 0x11 /* [RFC4893] */ #define BA_AS4_AGGREGATOR 0x12 +/* Protocol-specififc rte flags */ +#define PFR_BGP_ASCACHE PFR_FLAG1 /* 0x01 Set if bgp.first_as rte field is valid */ +/* Protocol-specific rte comparison flags */ +#define BGP_CFLAG_SKIPMED 0x01 /* Skip MED when comparing rte from different neigbour ASNs */ + /* BGP connection states */ #define BS_IDLE 0 diff --git a/proto/ospf/ospf.c b/proto/ospf/ospf.c index ce7ad37..12b209f 100644 --- a/proto/ospf/ospf.c +++ b/proto/ospf/ospf.c @@ -105,7 +105,7 @@ static int ospf_reload_routes(struct proto *p); static void ospf_rt_notify(struct proto *p, struct rtable *table UNUSED, net * n, rte * new, rte * old UNUSED, ea_list * attrs); -static int ospf_rte_better(struct rte *new, struct rte *old); +static int ospf_rte_better(struct rte *new, struct rte *old, u32 flags UNUSED); static int ospf_rte_same(struct rte *new, struct rte *old); static void ospf_disp(timer *timer); @@ -314,7 +314,7 @@ ospf_init(struct proto_config *c) /* If new is better return 1 */ static int -ospf_rte_better(struct rte *new, struct rte *old) +ospf_rte_better(struct rte *new, struct rte *old, u32 flags UNUSED) { if (new->u.ospf.metric1 == LSINFINITY) return 0; diff --git a/proto/rip/rip.c b/proto/rip/rip.c index 543aa30..6967c68 100644 --- a/proto/rip/rip.c +++ b/proto/rip/rip.c @@ -265,7 +265,7 @@ rip_rte_update_if_better(rtable *tab, net *net, struct proto *p, rte *new) rte *old; old = rte_find(net, p); - if (!old || p->rte_better(new, old) || + if (!old || p->rte_better(new, old, 0) || (ipa_equal(old->attrs->from, new->attrs->from) && (old->u.rip.metric != new->u.rip.metric)) ) rte_update(tab, net, p, p, new); @@ -909,7 +909,7 @@ rip_rte_same(struct rte *new, struct rte *old) static int -rip_rte_better(struct rte *new, struct rte *old) +rip_rte_better(struct rte *new, struct rte *old, u32 flags UNUSED) { struct proto *p = new->attrs->proto; -- 1.7.3.2 --------------070507060406060400060500--

1 0

bird6 and protocol device
by Alexander Shikoff 17 Nov '11

17 Nov '11

Hello All, I've started to set up an IPv6 route-server and got first issue: [...] protocol device { primary "bge0" 2001:7f8:63::ffff/64; } [...] # ifconfig bge0 ifconfig bge0 bge0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500 options=8009b<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,LINKSTATE> ether 00:21:9b:fc:d5:e4 inet x.x.x.0 netmask 0xfffffe00 broadcast x.x.x.255 inet6 2001:7f8:63::ffff prefixlen 64 nd6 options=3<PERFORMNUD,ACCEPT_RTADV> media: Ethernet autoselect (1000baseT <full-duplex>) status: active bird6 fails to start with error: bird: /usr/local/etc/bird6.conf, line 92: Invalid prefix /usr/local/etc/rc.d/bird6: WARNING: failed to start bird6 What's wrong? Thanks in advance! -- MINO-RIPE

3 2

cf-lex.l:283: warning: implicit declaration of function 'close'
by Alexander V. Chernikov 10 Nov '11

10 Nov '11

Hello list! It seems we should add unistd.h header to cf-lex.l to eliminate warning introduced by merging 'include' patch.

2 1

[PATCH] protocol templates
by Alexander V. Chernikov 02 Nov '11

02 Nov '11

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello list! Patch permits templates generation via 'template' keyword. Protocols can inherit templates (and templates can inherit templates) via 'inherit' keyword. At the moment patch supports BGP only (however adding other protocols is trivial) E.g. template bgp rr_client_general { local as 65000; multihop; rr client; } template bgp rr_client_bb { inherit bgp_rr_client_general; rr cluster id 1.2.3.4; } protocol bgp rr_cl_ft { inherit rr_client_bb; neighbor 10.4.7.2 as 65000; } Review/comments are welcome -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.14 (FreeBSD) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk6rK5oACgkQwcJ4iSZ1q2lTrgCfa6qKEgDv312ixHuQSg0/cAzY 6+cAoIGtXS7xWRKj4wXXpXBdUXN+kUxF =k0LL -----END PGP SIGNATURE-----

4 5

BIRD <-> Quagga OSPF Compatibility
by Roman Hoog Antink 02 Nov '11

02 Nov '11

Hi there Quagga 0.99.18 fails to use routes learned from bird 1.3.4 via OSPF and point-to-point links. The routes might show up in quagga's routing table, but the nexthop will never be the host running bird despite generating those static routes. In the log of quagga's ospfd you will find these entries: OSPF: ospf_nexthop_calculation(): could not determine nexthop for link This issue was discovered and discussed earlier in [1], however the topology patch can not be applied any more to bird nowadays due to structural changes. The attached patch is an update of that old version and applies to bird 1.3.4. Let me quote from RFC 2328, section 16.1.1 [2], namely: "If the destination is a directly connected network, or a router which connects to the calculating router via a point-to-point interface, no next hop IP address is required." As I understand this sentence and its context, the link data field shall be ignored in "the first case" where "the parent vertex is the root". So I figure, this is quagga's fault (source code [3]). Looking at the source code of quagga, the circumvention of the issue is easier to do in bird than a proper fix in quagga. This patch is not meant to be merged into mainstream of bird. It is for the desperate that need to run bird and quagga together. References: [1] http://marc.info/?l=bird-users&m=118009334429272&w=2 [2] http://tools.ietf.org/html/rfc2328#page-161 [3] quagga 0.99.18, ospf_spf.c line 578 Cheers, Roman

4 10

IPv6 rip-only routing - no exchange of routes
by Goesta Smekal 02 Nov '11

02 Nov '11

Dear list, looking for a small routing daemon for dynamic IPv6 routing on embedded devices, I came across bird recently. Looks pretty much like what I need, but I must have missed something ... The goal is to have two tiny routers, linked to a shared LAN segment (at eth0 in my test). Each router advertises an IPv6 prefix to a local network (wireless in the end, just another ethernet segment on eth1 so far). Nodes on the separate networks should be able to talk to each other via the two routers. And if a third router (ore some more) joins in, the new segment(s) should automatically appear in the routers routing tables - dynamic routing ;-) I compiled bird 1.3.4 from the sources on two test hosts (Debian GNU/Linux - not the embedded stuff yet) using the commands: root@edge01:/usr/src/bird-1.3.4# ./configure --prefix=/usr/local --enable-ipv6 --with-protocols=static,radv,rip root@edge01:/usr/src/bird-1.3.4# make all && make install My first config looked like this: ---8<--- # bird.conf # # dynamic routing configuration # # v0.1 2011-10-21 GS (goesta(a)smekal.at) # global parameters log syslog all; debug protocols { routes, events, packets }; router id 192.168.0.34; ## routing protocols # device parameters protocol device { scan time 10; } # kernel routing table protocol kernel { scan time 10; persist yes; learn yes; import all; export all; } # router advertisements protocol radv { interface "eth1"; } # dynamic routing using RIPng protocol rip{ authentication md5; password "verysecure"; honor neighbor; interface "eth*"; } ---8<--- Bird started fine on both nodes (obviously the router id was different, but the password was the same) but the routes did not appear on the other router. Router advertisements work fine. After some re-thinking and re-reading the docs I tried to add 'primary' statements to the "protocol device" section, and played with authentication and other rip specific parameters. Eventually I captured RIPng (UDP port 520) packets on the network with empty routes and after adding this: ---8<--- protocol direct{ interface "*"; } ---8<--- the packets contain the two prefixes, LAN and local subnet, I expect the routers to exchange. However the routes do not show up and wireshark mentions "Unknown address family 8193" in the packet details section of the RIPng packets. Here is what syslog says: ---8<--- Oct 28 21:13:50 edge01 bird6: device1: Initializing Oct 28 21:13:50 edge01 bird6: direct1: Initializing Oct 28 21:13:50 edge01 bird6: kernel1: Initializing Oct 28 21:13:50 edge01 bird6: radv1: Initializing Oct 28 21:13:50 edge01 bird6: rip1: Initializing Oct 28 21:13:50 edge01 bird6: device1: Starting Oct 28 21:13:50 edge01 bird6: device1: Scanning interfaces Oct 28 21:13:50 edge01 bird6: device1: State changed to feed Oct 28 21:13:50 edge01 bird6: direct1: Starting Oct 28 21:13:50 edge01 bird6: direct1: State changed to feed Oct 28 21:13:50 edge01 bird6: kernel1: Starting Oct 28 21:13:50 edge01 bird6: kernel1: State changed to feed Oct 28 21:13:50 edge01 bird6: radv1: Starting Oct 28 21:13:50 edge01 bird6: radv1: State changed to feed Oct 28 21:13:50 edge01 bird6: rip1: Starting Oct 28 21:13:50 edge01 bird6: rip1: Listening on (dummy), port 520, mode broadcast (::) Oct 28 21:13:50 edge01 bird6: rip1: State changed to feed Oct 28 21:13:50 edge01 bird6: Started Oct 28 21:13:50 edge01 bird6: device1: State changed to up Oct 28 21:13:50 edge01 bird6: direct1 < primary address ::1/128 on interface lo added Oct 28 21:13:50 edge01 bird6: direct1 < primary address 2001:aaaa:bbbb:ccb8::/64 on interface eth0 added Oct 28 21:13:50 edge01 bird6: direct1 > added [best] 2001:aaaa:bbbb:ccb8::/64 dev eth0 Oct 28 21:13:50 edge01 bird6: direct1 < secondary address fe80::/64 on interface eth0 added Oct 28 21:13:50 edge01 bird6: direct1 < primary address 2001:aaaa:bbbb:ccee::/64 on interface eth1 added Oct 28 21:13:50 edge01 bird6: direct1 > added [best] 2001:aaaa:bbbb:ccee::/64 dev eth1 Oct 28 21:13:50 edge01 bird6: direct1 < secondary address fe80::/64 on interface eth1 added Oct 28 21:13:50 edge01 bird6: direct1: State changed to up Oct 28 21:13:50 edge01 bird6: kernel1: Connected to table master Oct 28 21:13:50 edge01 bird6: kernel1 < rejected by protocol 2001:aaaa:bbbb:ccee::/64 dev eth1 Oct 28 21:13:50 edge01 bird6: kernel1 < rejected by protocol 2001:aaaa:bbbb:ccb8::/64 dev eth0 Oct 28 21:13:50 edge01 bird6: kernel1: State changed to up Oct 28 21:13:50 edge01 bird6: radv1 < interface lo goes up Oct 28 21:13:50 edge01 bird6: radv1 < primary address ::1/128 on interface lo added Oct 28 21:13:50 edge01 bird6: radv1 < interface eth0 goes up Oct 28 21:13:50 edge01 bird6: radv1 < primary address 2001:aaaa:bbbb:ccb8::/64 on interface eth0 added Oct 28 21:13:50 edge01 bird6: radv1 < secondary address fe80::/64 on interface eth0 added Oct 28 21:13:50 edge01 bird6: radv1 < interface eth1 goes up Oct 28 21:13:50 edge01 bird6: radv1: Adding interface eth1 Oct 28 21:13:50 edge01 bird6: radv1 < primary address 2001:aaaa:bbbb:ccee::/64 on interface eth1 added Oct 28 21:13:50 edge01 bird6: radv1 < secondary address fe80::/64 on interface eth1 added Oct 28 21:13:50 edge01 bird6: radv1: State changed to up Oct 28 21:13:50 edge01 bird6: rip1: Connected to table master Oct 28 21:13:50 edge01 bird6: rip1 < interface lo goes up Oct 28 21:13:50 edge01 bird6: rip1 < interface eth0 goes up Oct 28 21:13:50 edge01 bird6: rip1 < interface eth1 goes up Oct 28 21:13:50 edge01 bird6: rip1 < added 2001:aaaa:bbbb:ccee::/64 dev eth1 Oct 28 21:13:50 edge01 bird6: rip1 < added 2001:aaaa:bbbb:ccb8::/64 dev eth0 Oct 28 21:13:50 edge01 bird6: rip1: State changed to up Oct 28 21:13:50 edge01 bird6: kernel1: Scanning routing table Oct 28 21:13:50 edge01 bird6: kernel1: ::/0: [alien] created Oct 28 21:13:50 edge01 bird6: kernel1: Pruning table master Oct 28 21:13:50 edge01 bird6: kernel1: Pruning inherited routes Oct 28 21:13:51 edge01 bird6: kernel1 > added [best] ::/0 via fe80::20c:29ff:fe04:48bc on eth0 Oct 28 21:13:51 edge01 bird6: kernel1 < rejected by protocol ::/0 via fe80::20c:29ff:fe04:48bc on eth0 Oct 28 21:13:51 edge01 bird6: rip1 < added ::/0 via fe80::20c:29ff:fe04:48bc on eth0 Oct 28 21:13:51 edge01 bird6: rip1: Listening on eth1, port 520, mode multicast (ff02::9) Oct 28 21:13:51 edge01 bird6: rip1: Listening on eth0, port 520, mode multicast (ff02::9) Oct 28 21:13:51 edge01 bird6: radv1: Event Init on eth1 Oct 28 21:13:51 edge01 bird6: radv1: Timer fired on eth1 Oct 28 21:13:51 edge01 bird6: radv1: Sending RA via eth1 Oct 28 21:14:00 edge01 bird6: device1: Scanning interfaces Oct 28 21:14:00 edge01 bird6: rip1: Broadcasting routing table to eth0 Oct 28 21:14:00 edge01 bird6: rip1: Broadcasting routing table to eth1 Oct 28 21:14:00 edge01 bird6: kernel1: Scanning routing table Oct 28 21:14:00 edge01 bird6: kernel1: ::/0: [alien] seen Oct 28 21:14:00 edge01 bird6: kernel1: Pruning table master Oct 28 21:14:00 edge01 bird6: kernel1: Pruning inherited routes Oct 28 21:14:06 edge01 bird6: radv1: Timer fired on eth1 Oct 28 21:14:06 edge01 bird6: radv1: Sending RA via eth1 ---8<--- This is where the startup sequence is over, the last few lines repeat in a ten second interval (obviously, since this is configured in protocols "device" and "kernel" ... The configuration now looks like this: ---8<--- # bird.conf # # dynamic routing configuration # # v0.1 2011-10-21 GS (goesta(a)smekal.at) # global parameters log syslog all; debug protocols all; #debug protocols { routes, events, packets }; router id 192.168.0.34; ## routing protocols # device parameters protocol device { scan time 10; primary "eth0" 2001:aaaa:bbbb:ccb8::1:1:1; primary "eth1" 2001:aaaa:bbbb:ccee::1:1:1; } protocol direct{ interface "*"; } # kernel routing table protocol kernel { scan time 10; persist no; learn yes; import all; export all; } # router advertisements protocol radv { interface "eth1"; } #protocol static{ # export all; # route 2001:aaaa:bbbb:ccee::/64 via "eth1"; # route ::/0 via 2001:aaaa:bbbb:ccb8:20c:29ff:fe04:48bc; #} # dynamic routing using RIPng protocol rip{ authentication none; password "verysecure"; honor always; import all; export all; interface "eth*"; } ---8<--- As you can see, I also fiddled around with static routes, which did no good either. There must be something missing somewhere, but I cannot figure it out ... Can anybody help? thanks in advance, Goesta -- #!/usr/bin/perl foreach $c (split(/ /,"47 6f 65 73 74 61 20 53 6d 65 6b 61 6c 0d 0a")) { print pack("C", hex($c));}

4 6