number of networks, number of routes before and after filtering). If
you use <cf/count/ instead, only the statistics will be printed.
- <tag>enable|disable|restart <m/name/|"<m/pattern/"|all</tag>
- Enable, disable or restart a given protocol instance, instances matching the <cf><m/pattern/</cf> or <cf/all/ instances.
-
<tag>configure [soft] ["<m/config file/"]</tag>
Reload configuration from a given file. BIRD will smoothly
switch itself to the new configuration, protocols are
but new routes would be processed according to the new
filters.
+ <tag>enable|disable|restart <m/name/|"<m/pattern/"|all</tag>
+ Enable, disable or restart a given protocol instance, instances matching the <cf><m/pattern/</cf> or <cf/all/ instances.
+
<tag/down/
Shut BIRD down.
attributes to be transparent (for example does not prepend its AS number to
AS PATH attribute and keep MED attribute). Default: disabled.
+ <tag>enable route refresh <m/switch/</tag> When BGP speaker
+ changes its import filter, it has to re-examine all routes
+ received from its neighbor against the new filter. As these
+ routes might not be available, there is a BGP protocol
+ extension Route Refresh (specified in RFC 2918) that allows
+ BGP speaker to request re-advertisment of all routes from its
+ neighbor. This option specifies whether BIRD advertises this
+ capability and accepts such requests. Even when disabled, BIRD
+ can send route refresh requests. Default: on.
+
<tag>enable as4 <m/switch/</tag> BGP protocol was designed to use 2B AS numbers
and was extended later to allow 4B AS number. BIRD supports 4B AS extension,
but by disabling this option it can be persuaded not to advertise it and
CF_KEYWORDS(PASSWORD, FROM, PASSIVE, TO, ID, EVENTS, PACKETS, PROTOCOLS, INTERFACES)
CF_KEYWORDS(PRIMARY, STATS, COUNT, FOR, COMMANDS, PREEXPORT, GENERATE)
CF_KEYWORDS(LISTEN, BGP, V6ONLY, ADDRESS, PORT, PASSWORDS, DESCRIPTION)
+CF_KEYWORDS(RELOAD, REFEED)
CF_ENUM(T_ENUM_RTS, RTS_, DUMMY, STATIC, INHERIT, DEVICE, STATIC_DEVICE, REDIRECT,
RIP, OSPF, OSPF_IA, OSPF_EXT1, OSPF_EXT2, BGP, PIPE)
{ proto_xxable($2, 1); } ;
CF_CLI(RESTART, proto_patt, <protocol> | \"<pattern>\" | all, [[Restart protocol]])
{ proto_xxable($2, 2); } ;
+CF_CLI(RELOAD, proto_patt, <protocol> | \"<pattern>\" | all, [[Reload protocol]])
+{ proto_xxable($2, 3); } ;
+CF_CLI(REFEED, proto_patt, <protocol> | \"<pattern>\" | all, [[Refeed protocol BROKEN]])
+{ proto_xxable($2, 4); } ;
+
+
CF_CLI_HELP(DEBUG, ..., [[Control protocol debugging]])
CF_CLI(DEBUG, proto_patt debug_mask, (<protocol> | <pattern> | all) (all | off | { states | routes | filters | events | packets }), [[Control protocol debugging]])
rem_node(&p->n);
switch (p->core_state)
{
+ case FS_HUNGRY:
+ l = &inactive_proto_list;
+ break;
+ case FS_FEEDING:
case FS_HAPPY:
l = &active_proto_list;
break;
l = &flush_proto_list;
break;
default:
- l = &inactive_proto_list;
+ ASSERT(0);
}
proto_enqueue(l, p);
}
}
static void
-proto_feed(void *P)
+proto_feed_initial(void *P)
{
struct proto *p = P;
}
static void
-proto_schedule_feed(struct proto *p)
+proto_schedule_feed(struct proto *p, int initial)
{
DBG("%s: Scheduling meal\n", p->name);
p->core_state = FS_FEEDING;
proto_relink(p);
- p->attn->hook = proto_feed;
+ p->attn->hook = initial ? proto_feed_initial : proto_feed_more;
ev_schedule(p->attn);
}
+/**
+ * proto_request_feeding - request feeding routes to the protocol
+ * @p: given protocol
+ *
+ * Sometimes it is needed to send again all routes to the
+ * protocol. This is called feeding and can be requested by this
+ * function. This would cause protocol core state transition
+ * to FS_FEEDING (during feeding) and when completed, it will
+ * switch back to FS_HAPPY. This function can be called even
+ * when feeding is already running, in that case it is restarted.
+ */
+void
+proto_request_feeding(struct proto *p)
+{
+ ASSERT(p->proto_state == PS_UP);
+
+ /* If we are already feeding, we want to restart it */
+ if (p->core_state == FS_FEEDING)
+ {
+ /* Unless feeding is in initial state */
+ if (p->attn->hook == proto_feed_initial)
+ return;
+
+ rt_feed_baby_abort(p);
+ }
+
+ proto_schedule_feed(p, 0);
+}
+
/**
* proto_notify_state - notify core about protocol state change
* @p: protocol the state of which has changed
case PS_UP:
ASSERT(ops == PS_DOWN || ops == PS_START);
ASSERT(cs == FS_HUNGRY);
- proto_schedule_feed(p);
+ proto_schedule_feed(p, 1);
break;
case PS_STOP:
if ((cs = FS_FEEDING) || (cs == FS_HAPPY))
{
cli_msg(-9, "%s: disabled", p->name);
p->disabled = 1;
+ proto_rethink_goal(p);
}
break;
case 1:
{
cli_msg(-11, "%s: enabled", p->name);
p->disabled = 0;
+ proto_rethink_goal(p);
}
break;
case 2:
p->disabled = 1;
proto_rethink_goal(p);
p->disabled = 0;
+ proto_rethink_goal(p);
cli_msg(-12, "%s: restarted", p->name);
}
break;
+ case 3:
+ // FIXME change msg number
+ if (p->disabled)
+ cli_msg(-8, "%s: already disabled", p->name);
+ else if (p->reload_routes && p->reload_routes(p))
+ cli_msg(-12, "%s: reloading", p->name);
+ else
+ cli_msg(-12, "%s: reload failed", p->name);
+ break;
+ case 4:
+ // FIXME change msg number
+ if (p->disabled)
+ cli_msg(-8, "%s: already disabled", p->name);
+ else
+ {
+ proto_request_feeding(p);
+ cli_msg(-12, "%s: reexport failed", p->name);
+ }
+ break;
+
default:
ASSERT(0);
}
- proto_rethink_goal(p);
}
WALK_PROTO_LIST_END;
if (!cnt)
* It can construct a new rte, add private attributes and
* decide whether the route shall be imported: 1=yes, -1=no,
* 0=process it through the import filter set by the user.
+ * reload_routes Request protocol to reload all its routes to the core
+ * (using rte_update()). Returns: 0=reload cannot be done,
+ * 1= reload is scheduled and will happen (asynchronously).
*/
void (*if_notify)(struct proto *, unsigned flags, struct iface *i);
struct ea_list *(*make_tmp_attrs)(struct rte *rt, struct linpool *pool);
void (*store_tmp_attrs)(struct rte *rt, struct ea_list *attrs);
int (*import_control)(struct proto *, struct rte **rt, struct ea_list **attrs, struct linpool *pool);
+ int (*reload_routes)(struct proto *);
/*
* Routing entry hooks (called only for rte's belonging to this protocol):
void *proto_new(struct proto_config *, unsigned size);
void *proto_config_new(struct protocol *, unsigned size);
+void proto_request_feeding(struct proto *p);
void proto_show(struct symbol *, int);
struct proto *proto_get_named(struct symbol *, struct protocol *);
void proto_xxable(char *, int);
* HUNGRY/DOWN --> HUNGRY/START --> HUNGRY/UP -->
* FEEDING/UP --> HAPPY/UP --> FLUSHING/STOP|DOWN -->
* HUNGRY/STOP|DOWN --> HUNGRY/DOWN
+ *
+ * Sometimes, protocol might switch from HAPPY/UP to FEEDING/UP
+ * if it wants to refeed the routes (for example BGP does so
+ * as a result of received ROUTE-REFRESH request).
*/
#define FS_HUNGRY 0
else
p->stats.exp_withdraws_received++;
+ /* This is a tricky part - we don't know whether route 'old' was
+ exported to protocol 'p' or was filtered by the export filter.
+ We try tu run the export filter to know this to have a correct
+ value in 'old' argument of rt_update (and proper filter value)
+
+ FIXME - this is broken because 'configure soft' may change
+ filters but keep routes */
+
if (old)
{
if (p->out_filter == FILTER_REJECT)
}
}
+ /* FIXME - This is broken because of incorrect 'old' value (see above) */
if (!new && !old)
return;
ok = 0;
else if (!ic && d->export_mode > 1)
{
+ /* FIXME - this shows what should be exported according
+ to current filters, but not what was really exported.
+ 'configure soft' command may change the export filter
+ and do not update routes */
+
if (p1->out_filter == FILTER_REJECT ||
p1->out_filter && f_run(p1->out_filter, &e, &tmpa, rte_update_pool, FF_FORCE_TMPATTR) > F_ACCEPT)
ok = 0;
}
}
+static int
+bgp_reload_routes(struct proto *P)
+{
+ struct bgp_proto *p = (struct bgp_proto *) P;
+ if (!p->conn || !p->conn->peer_refresh_support)
+ return 0;
+
+ bgp_schedule_packet(p->conn, PKT_ROUTE_REFRESH);
+ return 1;
+}
+
static void
bgp_start_locked(struct object_lock *lock)
{
P->rte_better = bgp_rte_better;
P->import_control = bgp_import_control;
P->neigh_notify = bgp_neigh_notify;
+ P->reload_routes = bgp_reload_routes;
p->cf = c;
p->local_as = c->local_as;
p->remote_as = c->remote_as;
u32 default_local_pref; /* Default value for LOCAL_PREF attribute */
u32 default_med; /* Default value for MULTI_EXIT_DISC attribute */
int capabilities; /* Enable capability handshake [RFC3392] */
+ int enable_refresh; /* Enable local support for route refresh [RFC2918] */
int enable_as4; /* Enable local support for 4B AS numbers [RFC4893] */
u32 rr_cluster_id; /* Route reflector cluster ID, if different from local ID */
int rr_client; /* Whether neighbor is RR client of me */
int start_state; /* protocol start_state snapshot when connection established */
int want_as4_support; /* Connection tries to establish AS4 session */
int peer_as4_support; /* Peer supports 4B AS numbers [RFC4893] */
+ int peer_refresh_support; /* Peer supports route refresh [RFC2918] */
unsigned hold_time, keepalive_time; /* Times calculated from my and neighbor's requirements */
};
#define PKT_UPDATE 0x02
#define PKT_NOTIFICATION 0x03
#define PKT_KEEPALIVE 0x04
+#define PKT_ROUTE_REFRESH 0x05
#define PKT_SCHEDULE_CLOSE 0x1f /* Used internally to schedule socket close */
/* Attributes */
BGP_ATOMIC_AGGR, BGP_AGGREGATOR, BGP_COMMUNITY, SOURCE, ADDRESS,
PASSWORD, RR, RS, CLIENT, CLUSTER, ID, AS4, ADVERTISE, IPV4,
CAPABILITIES, LIMIT, PASSIVE, PREFER, OLDER, MISSING, LLADDR,
- DROP, IGNORE)
+ DROP, IGNORE, ROUTE, REFRESH)
CF_GRAMMAR
BGP_CFG->error_amnesia_time = 300;
BGP_CFG->error_delay_time_min = 60;
BGP_CFG->error_delay_time_max = 300;
+ BGP_CFG->enable_refresh = 1;
BGP_CFG->enable_as4 = bgp_as4_support;
BGP_CFG->capabilities = 2;
BGP_CFG->advertise_ipv4 = 1;
| bgp_proto ERROR FORGET TIME expr ';' { BGP_CFG->error_amnesia_time = $5; }
| bgp_proto ERROR WAIT TIME expr ',' expr ';' { BGP_CFG->error_delay_time_min = $5; BGP_CFG->error_delay_time_max = $7; }
| bgp_proto DISABLE AFTER ERROR bool ';' { BGP_CFG->disable_after_error = $5; }
+ | bgp_proto ENABLE ROUTE REFRESH bool ';' { BGP_CFG->enable_refresh = $5; }
| bgp_proto ENABLE AS4 bool ';' { BGP_CFG->enable_as4 = $4; }
| bgp_proto CAPABILITIES bool ';' { BGP_CFG->capabilities = $3; }
| bgp_proto ADVERTISE IPV4 bool ';' { BGP_CFG->advertise_ipv4 = $4; }
}
#endif
+static byte *
+bgp_put_cap_rr(struct bgp_conn *conn UNUSED, byte *buf)
+{
+ *buf++ = 2; /* Capability 2: Support for route refresh */
+ *buf++ = 0; /* Capability data length */
+ return buf;
+}
+
static byte *
bgp_put_cap_as4(struct bgp_conn *conn, byte *buf)
{
cap = bgp_put_cap_ipv6(conn, cap);
#endif
+ if (p->cf->enable_refresh)
+ cap = bgp_put_cap_rr(conn, cap);
+
if (conn->want_as4_support)
cap = bgp_put_cap_as4(conn, cap);
#endif
+static byte *
+bgp_create_route_refresh(struct bgp_conn *conn, byte *buf)
+{
+ struct bgp_proto *p = conn->bgp;
+ BGP_TRACE(D_PACKETS, "Sending ROUTE-REFRESH");
+
+#ifdef IPV6
+ *buf++ = 0; /* AFI IPv6 */
+ *buf++ = BGP_AF_IPV6;
+#else
+ *buf++ = 0; /* AFI IPv4 */
+ *buf++ = BGP_AF_IPV4;
+#endif
+ *buf++ = 0; /* RFU */
+ *buf++ = 1; /* and SAFI 1 */
+ return buf;
+}
+
static void
bgp_create_header(byte *buf, unsigned int len, unsigned int type)
{
type = PKT_OPEN;
end = bgp_create_open(conn, pkt);
}
+ else if (s & (1 << PKT_ROUTE_REFRESH))
+ {
+ s &= ~(1 << PKT_ROUTE_REFRESH);
+ type = PKT_ROUTE_REFRESH;
+ end = bgp_create_route_refresh(conn, pkt);
+ }
else if (s & (1 << PKT_UPDATE))
{
end = bgp_create_update(conn, pkt);
switch (opt[0])
{
+ case 2:
+ if (cl != 0)
+ goto err;
+ conn->peer_refresh_support = 1;
+ break;
case 65:
if (cl != 4)
goto err;
}
}
+static void
+bgp_rx_route_refresh(struct bgp_conn *conn, byte *pkt, int len)
+{
+ struct bgp_proto *p = conn->bgp;
+
+ BGP_TRACE(D_PACKETS, "Got ROUTE-REFRESH");
+
+ if (conn->state != BS_ESTABLISHED)
+ { bgp_error(conn, 5, 0, NULL, 0); return; }
+
+ if (!p->cf->enable_refresh)
+ { bgp_error(conn, 1, 3, pkt+18, 1); return; }
+
+ if (len != (BGP_HEADER_LENGTH + 4))
+ { bgp_error(conn, 1, 2, pkt+16, 2); return; }
+
+ /* FIXME - we ignore AFI/SAFI values, as we support
+ just one value and even an error code for an invalid
+ request is not defined */
+
+ proto_request_feeding(&p->p);
+}
+
+
/**
* bgp_rx_packet - handle a received packet
* @conn: BGP connection
case PKT_UPDATE: return bgp_rx_update(conn, pkt, len);
case PKT_NOTIFICATION: return bgp_rx_notification(conn, pkt, len);
case PKT_KEEPALIVE: return bgp_rx_keepalive(conn);
+ case PKT_ROUTE_REFRESH: return bgp_rx_route_refresh(conn, pkt, len);
default: bgp_error(conn, 1, 3, pkt+18, 1);
}
}
projects / thirdparty / bird.git / commitdiff
