- Bird-users - Mailing Lists

Another update to the BGPsec BIRD implementation
by Michael Baer 10 May '16

10 May '16

Hi all, I wanted to announce an updated versions of the BGPsec supporting code using BIRD. The new BGPsec code is version 0.7 and is based on BIRD 1.5.0. It is available as a source tarballs at a new domain: https://securerouting.net/bird The main changes from the last release is that it has been updated to match the most recent BGPsec specification, draft-ietf-sidr-bgpsec-protocol-15, and it has been updated to version 1.5.0 of the BIRD code. This is an ongoing project following the IETF's SIDR Working Group's RPKI/BGPsec work: http://datatracker.ietf.org/wg/sidr/charter/ http://datatracker.ietf.org/doc/draft-ietf-sidr-bgpsec-protocol/ The code is still at the testing stage and should not be use used for production services. If any one wants to test the code, please feel free to email me any questions or bug reports. Thanks, Mike -- Michael Baer PARSONS baerm(a)tislabs.com

1 0

[PATCH] netlink: update struct msghdr
by Stijn Tintel 10 May '16

10 May '16

The netlink code assumes an order for the members of struct msghdr. This breaks recvmsg and sendmsg with musl libc on mips64. Fix this by using designated initializers instead. Signed-off-by: Stijn Tintel <stijn(a)linux-ipv6.be> --- sysdep/linux/netlink.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/sysdep/linux/netlink.c b/sysdep/linux/netlink.c index 1ffdff0..b42e7b6 100644 --- a/sysdep/linux/netlink.c +++ b/sysdep/linux/netlink.c @@ -125,7 +125,12 @@ nl_get_reply(struct nl_sock *nl) { struct iovec iov = { nl->rx_buffer, NL_RX_SIZE }; struct sockaddr_nl sa; - struct msghdr m = { (struct sockaddr *) &sa, sizeof(sa), &iov, 1, NULL, 0, 0 }; + struct msghdr m = { + .msg_name = &sa, + .msg_namelen = sizeof(sa), + .msg_iov = &iov, + .msg_iovlen = 1, + }; int x = recvmsg(nl->fd, &m, 0); if (x < 0) die("nl_get_reply: %m"); @@ -1231,7 +1236,12 @@ nl_async_hook(sock *sk, int size UNUSED) { struct iovec iov = { nl_async_rx_buffer, NL_RX_SIZE }; struct sockaddr_nl sa; - struct msghdr m = { (struct sockaddr *) &sa, sizeof(sa), &iov, 1, NULL, 0, 0 }; + struct msghdr m = { + .msg_name = &sa, + .msg_namelen = sizeof(sa), + .msg_iov = &iov, + .msg_iovlen = 1, + }; struct nlmsghdr *h; int x; uint len; -- 2.7.3

2 1

Displaying all route attributes
by Baptiste Jonglez 03 May '16

03 May '16

Hi, Is it possible to display all attributes attached to a route? "show route all" displays only a subset of the attributes, and the names are different (e.g. "BGP.as_path" in "show route all" corresponds to the bgp_path attribute). Thanks, Baptiste

2 1

systemd unit and ospf script
by Rohrmann Sascha 03 May '16

03 May '16

Hi, I recently noticed, that there isn't a systemd unit for bird. That's why I want to share mine. You could add them into the bird package @ondrej Further information: you need to put this into /etc/systemd/system Furthermore I wrote a script in perl which basicly activates or disables the originating of the default-route into the ospf area relating to incoming BGP routes. The needed perl module is from this page: https://github.com/stephank/nagios-bird You need to download birdctl.pm and put it in the same folder as the script. Next you need to adjust the ENV variable to your folder. As a little extra I created a systemd unit for my script. You need to adjust the path here, too. Greetings, Sascha.

1 0

[PATCH 4/7] Babel: Documentation updates.
by Toke Høiland-Jørgensen 03 May '16

03 May '16

This updates the documentation to correctly mention Babel when protocols are listed, and adds examples and route attribute documentation to the Babel section of the docs. Signed-off-by: Toke Høiland-Jørgensen <toke(a)toke.dk> --- doc/bird.sgml | 33 ++++++++++++++++++++++++++++++--- nest/config.Y | 2 +- nest/rt-attr.c | 2 +- 3 files changed, 32 insertions(+), 5 deletions(-) diff --git a/doc/bird.sgml b/doc/bird.sgml index a0b45f4..71fcb1c 100644 --- a/doc/bird.sgml +++ b/doc/bird.sgml @@ -610,8 +610,8 @@ agreement"). options, in that case for given interface the first matching interface option is used. - This option is allowed in BFD, Direct, OSPF, RAdv and RIP protocols, but - in OSPF protocol it is used in the <cf/area/ subsection. + This option is allowed in Babel, BFD, Direct, OSPF, RAdv and RIP + protocols, but in OSPF protocol it is used in the <cf/area/ subsection. Default: none. @@ -1312,7 +1312,7 @@ clist for most purposes. <cf/RTS_DUMMY/, <cf/RTS_STATIC/, <cf/RTS_INHERIT/, <cf/RTS_DEVICE/, <cf/RTS_STATIC_DEVICE/, <cf/RTS_REDIRECT/, <cf/RTS_RIP/, <cf/RTS_OSPF/, <cf/RTS_OSPF_IA/, <cf/RTS_OSPF_EXT1/, <cf/RTS_OSPF_EXT2/, <cf/RTS_BGP/, - <cf/RTS_PIPE/. + <cf/RTS_PIPE/, <cf/RTS_BABEL/. <tag><m/enum/ cast</tag> Route type (Currently <cf/RTC_UNICAST/ for normal routes, @@ -1475,6 +1475,33 @@ protocol babel [<name>] { yes. </descrip> +<sect1>Attributes + +<p>Babel defines just one attribute: the internal babel metric of the route. It +is exposed as the <cf/babel_metric/ attribute and has range from 1 to infinity +(65535). + +<sect1>Example + +<p><code> +protocol babel { + interface "eth*" { + type wired; + }; + interface "wlan0", "wlan1" { + type wireless; + hello interval 1; + rxcost 512; + }; + interface "tap0"; + + # This matches the default of babeld: redistribute all addresses + # configured on local interfaces, plus re-distribute all routes received + # from other babel peers. + export where (source = RTS_DEVICE) || (source = RTS_BABEL); +} +</code> + <sect><label id="sect-bfd">BFD diff --git a/nest/config.Y b/nest/config.Y index 4566971..ca70564 100644 --- a/nest/config.Y +++ b/nest/config.Y @@ -63,7 +63,7 @@ CF_KEYWORDS(RELOAD, IN, OUT, MRTDUMP, MESSAGES, RESTRICT, MEMORY, IGP_METRIC, CL CF_KEYWORDS(GRACEFUL, RESTART, WAIT, MAX, FLUSH, AS) CF_ENUM(T_ENUM_RTS, RTS_, DUMMY, STATIC, INHERIT, DEVICE, STATIC_DEVICE, REDIRECT, - RIP, OSPF, OSPF_IA, OSPF_EXT1, OSPF_EXT2, BGP, PIPE) + RIP, OSPF, OSPF_IA, OSPF_EXT1, OSPF_EXT2, BGP, PIPE, BABEL) CF_ENUM(T_ENUM_SCOPE, SCOPE_, HOST, LINK, SITE, ORGANIZATION, UNIVERSE, UNDEFINED) CF_ENUM(T_ENUM_RTC, RTC_, UNICAST, BROADCAST, MULTICAST, ANYCAST) CF_ENUM(T_ENUM_RTD, RTD_, ROUTER, DEVICE, BLACKHOLE, UNREACHABLE, PROHIBIT, MULTIPATH) diff --git a/nest/rt-attr.c b/nest/rt-attr.c index 7fa05d6..b57bcd5 100644 --- a/nest/rt-attr.c +++ b/nest/rt-attr.c @@ -1160,7 +1160,7 @@ rta_dump(rta *a) static char *rts[] = { "RTS_DUMMY", "RTS_STATIC", "RTS_INHERIT", "RTS_DEVICE", "RTS_STAT_DEV", "RTS_REDIR", "RTS_RIP", "RTS_OSPF", "RTS_OSPF_IA", "RTS_OSPF_EXT1", - "RTS_OSPF_EXT2", "RTS_BGP" }; + "RTS_OSPF_EXT2", "RTS_BGP", "RTS_PIPE", "RTS_BABEL" }; static char *rtc[] = { "", " BC", " MC", " AC" }; static char *rtd[] = { "", " DEV", " HOLE", " UNREACH", " PROHIBIT" }; -- 2.8.0

2 3

Unable to access https://gitlab.labs.nic.cz
by David S. 03 May '16

03 May '16

Hi All, I'm from Indonesia, I'm unable to access https://gitlab.labs.nic.cz, I have tried to access the url from 3 different network provider and I get same result. Here is the telnet test result: $ telnet gitlab.labs.nic.cz 443 Trying 217.31.192.88... telnet: connect to address 217.31.192.88: Operation timed out Trying 2001:1488:ac15:ff70::88... telnet: connect to address 2001:1488:ac15:ff70::88: No route to host telnet: Unable to connect to remote host Does anyone have information about this? Thank you Best regards, David S. ------------------------------------------------ e. david(a)zeromail.us w. pnyet.web.id p. 087881216110

3 3

[PATCH 3/7] Babel: Rework handling of retractions.
by Toke Høiland-Jørgensen 02 May '16

02 May '16

An update with wildcard AE and infinite metric should be treated as a global retraction of all prefixes announced by that neighbour, per section 4.4.9 of the RFC. In addition, router ID and seqno in retraction updates should be ignored. This reworks the handling of retractions and adjusts the parser to handle all this correctly. Signed-off-by: Toke Høiland-Jørgensen <toke(a)toke.dk> --- proto/babel/babel.c | 72 +++++++++++++++++++++++++++++++++++---------------- proto/babel/packets.c | 22 +++++++++++----- 2 files changed, 65 insertions(+), 29 deletions(-) diff --git a/proto/babel/babel.c b/proto/babel/babel.c index 4b28f69..e982abb 100644 --- a/proto/babel/babel.c +++ b/proto/babel/babel.c @@ -1040,17 +1040,18 @@ babel_handle_update(union babel_msg *m, struct babel_iface *ifa) struct babel_proto *p = ifa->proto; struct babel_msg_update *msg = &m->update; - struct babel_neighbor *n; + struct babel_neighbor *nbr; struct babel_entry *e; struct babel_source *s; struct babel_route *r; + node *n; int feasible; TRACE(D_PACKETS, "Handling update for %I/%d with seqno %d metric %d", msg->prefix, msg->plen, msg->seqno, msg->metric); - n = babel_find_neighbor(ifa, msg->sender); - if (!n) + nbr = babel_find_neighbor(ifa, msg->sender); + if (!nbr) { DBG("Babel: Haven't heard from neighbor %I; ignoring update.\n", msg->sender); return; @@ -1095,27 +1096,54 @@ babel_handle_update(union babel_msg *m, struct babel_iface *ifa) * of the Interval value included in the update. */ - if (msg->metric == BABEL_INFINITY) - e = babel_find_entry(p, msg->prefix, msg->plen); - else - e = babel_get_entry(p, msg->prefix, msg->plen); + if (msg->metric == BABEL_INFINITY) /* Retraction */ + { + if (msg->ae == BABEL_AE_WILDCARD) + { + /* Special case: This is a retraction of all prefixes announced by this + neighbour (see second-to-last paragraph of section 4.4.9 in the + RFC). */ + WALK_LIST(n, nbr->routes) + { + r = SKIP_BACK(struct babel_route, neigh_route, n); + r->metric = BABEL_INFINITY; + babel_select_route(r->e); + } + } + else + { - if (!e) - return; + e = babel_find_entry(p, msg->prefix, msg->plen); + + if (!e) + return; + + r = babel_find_route(e, nbr); /* the route entry indexed by neighbour */ + if(!r) + return; + + r->metric = BABEL_INFINITY; + babel_select_route(e); + } + + return; /* Done with retractions */ + } + + e = babel_get_entry(p, msg->prefix, msg->plen); + r = babel_find_route(e, nbr); /* the route entry indexed by neighbour */ s = babel_find_source(e, msg->router_id); /* for feasibility */ - r = babel_find_route(e, n); /* the route entry indexed by neighbour */ feasible = babel_is_feasible(s, msg->seqno, msg->metric); if (!r) { - if (!feasible || (msg->metric == BABEL_INFINITY)) + if (!feasible) return; - r = babel_get_route(e, n); + r = babel_get_route(e, nbr); r->advert_metric = msg->metric; r->router_id = msg->router_id; - r->metric = babel_compute_metric(n, msg->metric); + r->metric = babel_compute_metric(nbr, msg->metric); r->next_hop = msg->next_hop; r->seqno = msg->seqno; } @@ -1126,24 +1154,22 @@ babel_handle_update(union babel_msg *m, struct babel_iface *ifa) babel_unicast_seqno_request(r); if (msg->router_id == r->router_id) return; - r->metric = BABEL_INFINITY; /* retraction */ + r->metric = BABEL_INFINITY; /* treat as retraction */ } else { /* Last paragraph above - update the entry */ r->advert_metric = msg->metric; - r->metric = babel_compute_metric(n, msg->metric); - r->router_id = msg->router_id; + r->metric = babel_compute_metric(nbr, msg->metric); r->next_hop = msg->next_hop; + + r->router_id = msg->router_id; r->seqno = msg->seqno; - if (msg->metric != BABEL_INFINITY) - { - r->expiry_interval = BABEL_ROUTE_EXPIRY_FACTOR(msg->interval); - r->expires = now + r->expiry_interval; - if (r->expiry_interval > BABEL_ROUTE_REFRESH_INTERVAL) - r->refresh_time = now + r->expiry_interval - BABEL_ROUTE_REFRESH_INTERVAL; - } + r->expiry_interval = BABEL_ROUTE_EXPIRY_FACTOR(msg->interval); + r->expires = now + r->expiry_interval; + if (r->expiry_interval > BABEL_ROUTE_REFRESH_INTERVAL) + r->refresh_time = now + r->expiry_interval - BABEL_ROUTE_REFRESH_INTERVAL; /* If the route is not feasible at this point, it means it is from another neighbour than the one currently selected; so send a unicast seqno diff --git a/proto/babel/packets.c b/proto/babel/packets.c index be47aa7..60b9cd3 100644 --- a/proto/babel/packets.c +++ b/proto/babel/packets.c @@ -484,6 +484,13 @@ babel_read_update(struct babel_tlv *hdr, union babel_msg *m, break; case BABEL_AE_IP4: + if (tlv->flags & BABEL_FLAG_ROUTER_ID) + { + /* spec doesn't specify how to treat updates with the router ID flag set + * for an IPv4 address. We consider them an error. */ + DBG("Received update with FLAG_ROUTER_ID set for an IPv4 address.\n"); + return PARSE_ERROR; + } /* TODO */ return PARSE_IGNORE; @@ -508,11 +515,6 @@ babel_read_update(struct babel_tlv *hdr, union babel_msg *m, state->def_ip6_prefix_seen = 1; } - if (tlv->flags & BABEL_FLAG_ROUTER_ID) - { - state->router_id = ((u64) _I2(msg->prefix)) << 32 | _I3(msg->prefix); - state->router_id_seen = 1; - } break; case BABEL_AE_IP6_LL: @@ -523,7 +525,15 @@ babel_read_update(struct babel_tlv *hdr, union babel_msg *m, return PARSE_IGNORE; } - if (!state->router_id_seen) + if (tlv->flags & BABEL_FLAG_ROUTER_ID) + { + state->router_id = ((u64) _I2(msg->prefix)) << 32 | _I3(msg->prefix); + state->router_id_seen = 1; + } + + /* Updates with no router ID are allowed for retractions (RFC section + * 4.4.9). */ + if (!state->router_id_seen && msg->metric != BABEL_INFINITY) { DBG("Babel: No router ID seen before update\n"); return PARSE_ERROR; -- 2.8.0

1 0

Release plans?
by Shaun Crampton 27 Apr '16

27 Apr '16

Hi, I was just wondering if the BIRD team had any plans to do a release any time soon (I think that would be 1.6?). We'd really like to have a release with BGP multi-path support. Regards, -Shaun

5 10

[mrtdump branch] Segfault when using a long filename for mrtdump
by Baptiste Jonglez 27 Apr '16

27 Apr '16

Hi, I have been experimenting with the mrtdump branch (commit cc4eee62) on Debian jessie. However, when using a moderately long filename, Bird crashes. For instance: birdc6 'mrtdump routes to "/bird/mrtdump/rib.ipv6.20160426.1209"' BIRD 1.5.0 ready. Connection closed by server. The logs are the following: bird6[8314]: Unable to open file "<too-long>" for MRT dump of table master systemd[1]: bird6.service: main process exited, code=killed, status=11/SEGV systemd[1]: Unit bird6.service entered failed state. kernel: bird6[8314]: segfault at 18 ip 00007fb05162a905 sp 00007ffe1194f5a0 error 4 in bird6[7fb051602000+76000] It looks like TM_DATETIME_BUFFER_SIZE, as used in tm_format_datetime(), is really too small (32 bytes). Also, there seems to be an issue with error handling (segfault when tm_format_datetime returns unexpected data). When defining a higher value for TM_DATETIME_BUFFER_SIZE, bird does not crash anymore, but there is certainly a better solution. Thanks, Baptiste

2 4

Using BIRD to inject a route into BGP
by Eric Vyncke (evyncke) 25 Apr '16

25 Apr '16

Mainly for lab experimentation, I would like to be able to dynamically inject several IPv6 routes into BGP. Creating a static route on the Linux running BIRD does not seem to grab the kernel route (even when doing "ip -6 route add ... proto [static|kernel]..." My config has "import all;export all" for kernel & device protocols + "learn on" What am I missing? Is it possible with BIRD without having to configure those routes in protocol static and reload BIRD? (which is a little too heavy for my environment) Thanks in advance -éric

4 5