http://blog.gmane.org/gmane.network.quagga.devel hourly 1 1901-01-01T00:00+00:00 http://gmane.org/img/gmane-25t.png http://gmane.org http://comments.gmane.org/gmane.network.quagga.devel/9399 <pre>Hi. On Receiving Ack with ospfd, one more check is necessary. When a ls-update packet with non-MaxAge LSA has appeared, the LSA is registerred on retransmit list. and if the LSA is overwritten with MaxAge &amp; same-seqnum, LSA on retransmit list is also replaced with MaxAge one. when an neighbor slowly receives non-MaxAge LSA and send back an Ack, current ospfd detects this is "same instance" and delete the MaxAge LSA from retrans list. This causes LSDB inconsistency. melody(15:00:47):~/work/quagga-0.99.20.1/ospfd$ diff -U10 ospf_packet.c ospf_packet.2.c --- ospf_packet.c2012-05-25 15:00:41.000000000 +0900 +++ ospf_packet.2.c2012-05-25 15:00:26.000000000 +0900 &lt; at &gt;&lt; at &gt; -2077,21 +2077,22 &lt; at &gt;&lt; at &gt; if (lsa-&gt;data-&gt;type &lt; OSPF_MIN_LSA || lsa-&gt;data-&gt;type &gt;= OSPF_MAX_LSA) { lsa-&gt;data = NULL; ospf_lsa_discard (lsa); continue; } lsr = ospf_ls_retransmit_lookup (nbr, lsa); - if (lsr != NULL &amp;&amp; lsr-&gt;data-&gt;ls_seqnum == lsa-&gt;data-&gt;ls_seqnum) + if (lsr != NULL &amp;&amp; lsr-&gt;data-&gt;ls_seqnum == lsa-&gt;data-&gt;ls_seqnum + &amp;&amp; (ntohs (lsr-&gt;data-&gt;ls_age) &gt;= OSPF_LSA_MAXAGE) == (ntohs (lsr-&gt;data-&gt;ls_age) &gt;= OSPF_LSA_MAXAGE)) { #ifdef HAVE_OPAQUE_LSA if (IS_OPAQUE_LSA (lsr-&gt;data-&gt;type)) ospf_opaque_ls_ack_received (nbr, lsr); #endif /* HAVE_OPAQUE_LSA */ ospf_ls_retransmit_delete (nbr, lsr); } lsa-&gt;data = NULL; melody(15:00:58):~/work/quagga-0.99.20.1/ospfd$ Thanks. -- kashima&lt; at &gt;jp.fujitsu.com </pre> KASHIMA Hiroaki 2012-05-25T06:15:49 http://comments.gmane.org/gmane.network.quagga.devel/9395 <pre>Hi , I am trying to cross compile quagga 0.99.18 routing suite for octeon . After setting "ac_cv_func_malloc_0_nonnull=yes ac_cv_func_realloc_0_nonnull=yes" in configure my cross compilation was successful . But when I tried to run the zebra daemon on the octeon board I got a segmentation fault. please suggest a way to go about this issue. Also please provide a solution to get statically linked quagga daemons as the -static flag is ineffective. Thanks &amp; Regards, Harini Gopalakrishnan ________________________________ The contents of this e-mail and any attachment(s) may contain confidential or privileged information for the intended recipient(s). Unintended recipients are prohibited from taking action on the basis of information in this e-mail and using or disseminating the information, and must notify the sender and delete it from their system. L&amp;T Infotech will not accept responsibility or liability for the accuracy or completeness of, or the presence of any virus or disabling code in this e-mail" </pre> Harini Gopalakrishnan 2012-05-24T09:53:58 http://comments.gmane.org/gmane.network.quagga.devel/9394 <pre>Hi! Here is a serie of patches adding AgentX support to Quagga. The first patch is not related to AgentX support. It just fixes Quagge compilation when using a separate build directory. The second patch removes some cruft left from a migration from UCD-SNMP (I think). This is just a small cleanup. The third patch makes use of "net-snmp-config" in place of linking to -lnetsnmp. It removes explicit link to -lcrypto. If needed, this will be added by "net-snmp-config". "net-snmp-config" exists since at least NetSNMP 4.2. The fourth patch separates SMUX specific stuff from generic SNMP stuff. smux.h now only contains the interface to SMUX for the daemons. Everything else has been moved to smux.c. The idea is for another SNMP implementation to provide the appropriate smux_* functions. snmp.c is a new file containing oid_* functions (they are independant of the SNMP implementation). The last patch adds AgentX support by providing the appropriate smux_* functions. This patch does not modify daemons. Because Net-SNMP hides its internals, it is not possible to disable AgentX support once enabled. This should not be a problem, because, like with SMUX, the commands are not exported to vtysh and therefore, I suppose there is no way to call "no agentx". Also, the event loop is modified (in thread.c). NetSNMP hides the appropriate file descriptors and timers. The documentation has not been updated. This should be easy enough once I am done with the other details. The big problem now is traps. Currently, I do not send any trap with AgentX. The SMUX implementation only sends SNMPv1 traps using the SMUX peer OID as an enterprise OID. I think this is not the right way to do it. Both OSPF-MIB (through OSPF-TRAP-MIB) and BGP-MIB defines notifications. A notification is some kind of SNMPv2 TRAP. Let me explain. With a SNMPv1 trap, you have: an enterprise OID, a generic trap code (always 6 in our case) and a specific trap code. With a notification, all those are replaced by a snmptrap OID. It is possible to transform SNMPv1 trap to SNMPv2 trap and vice-versa. However, if I just do this, I will continue to use inappropriate enterprise OID (the SMUX peer OID). I should use the appropriate notification objects, as defined in MIB. My first proposition was to change smux_trap() signature to include the appropriate enterprise OID (used to build snmp trap OID). Then, I will also update the SMUX implementation. There are two drawbacks: - daemons maintained outside of Quagga will also need to update (if they use traps) - people receiving traps will need to update their configuration to use the correct OID instead of the previous ones. Another proposition will be to just keep the SMUX peer OID. People may change it to the appropriate value (since I will not use it as a SMUX peer OID) to get correct snmp trap OID. For BGP and OSPF, this will work because all traps are rooted at the same OID. I can however provide a "smux_inform()" function allowing to override the enterprise OID. Any though? And any comment on the patches in general? </pre> Vincent Bernat 2012-05-24T08:56:30 http://comments.gmane.org/gmane.network.quagga.devel/9379 <pre>Hi list, under linux &gt; 2.6.18, quagga (all versions) would not remove the "directly connected" state routes when IP addresses were removed from interfaces in vtysh. My "fix" was to remove the block checking if the message has the same pid as the command interface. I hope this doesn't make zebra parse lots of irrelevant messages. The patch is attached to bug #721, where this is also discussed in greater detail. I'll also attach it to this mail. Best regards, Torsten </pre> Torsten Luettgert 2012-05-22T16:24:54 http://comments.gmane.org/gmane.network.quagga.devel/9378 <pre>Hi! Due to the serious problems[1] with ospfd in Quagga &gt;0.99.17 we've decided at Westermo to downgrade to 0.99.17 for our upcoming WeOS[2] 4.9.0 release. Contrary to earlier releases of WeOS I've decided to personally steal some corporate time to communicating our efforts and publicizing our patches at GitHub[3]. https://github.com/troglobit/quagga/tree/security/0.99.17 https://github.com/troglobit/quagga/tree/westermo/0.99.17 So far I've setup the repository, created a security/0.99.17 branch, with patches from the official security/0.99.20 tree, and added the Westermo patches (including Stephen Hemminger's Zebra route manager patch! :) to a westermo/0.99.17 tree. Before our own WeOS release I hope to add CVE fixes from Debian[4], at the very least CVE-2011-3325 and CVE-2011-3326, i.e., DSA-2316-1 see[5] for more details. Maybe someone else is interested in this effort. Iirc Joakim Tjernlund also use 0.99.17 as a baseline for Transmode ... Regards /Joachim [1] - http://lists.quagga.net/pipermail/quagga-dev/2012-May/009391.html [2] - http://www.westermo.com [3] - http://github.com/troglobit/quagga [4] - http://security-tracker.debian.org/tracker/source-package/quagga [5] - http://security-tracker.debian.org/tracker/DSA-2316-1 </pre> Joachim Nilsson 2012-05-21T18:16:31 http://comments.gmane.org/gmane.network.quagga.devel/9371 <pre>Hi! Implementation of OSPFv3-MIB in Quagga is incomplete and outdated: it uses an experimental MIB instead of the one from RFC 5643 and only ospfv3AreaLsdbTable is implemented. Is someone working on a more current implementation? </pre> Vincent Bernat 2012-05-17T09:03:30 http://comments.gmane.org/gmane.network.quagga.devel/9369 <pre>Hi, I was wondering if in the function - ospf6_route_remove_all() Are we not accessing freed memory (route) while trying to get the route next using ospf6_route_next(). Similar is the case in ospf6_intra_brouter_calculation() function. We are freeing the memory in ospf6_route_remove()--&gt; ospf6_route_unlock() and on return to get the next route entry we use the same memory freed memory again. Is it not an issue. Regards Narender </pre> narender 2012-05-16T14:32:55 http://comments.gmane.org/gmane.network.quagga.devel/9366 <pre>Is there an automated test suite available for quagga? I saw some mentions of tests with respect to the Open Source Routing group, but I didn't see anything about tests I could run myself. I do see that there's a tests/ directory, but it looks pretty sparse... is that all that's available or am I missing tests that are stored elsewhere? </pre> Terri Oda 2012-05-14T23:07:56 http://comments.gmane.org/gmane.network.quagga.devel/9365 <pre>Hi! We've seen OSPF issues when reconfiguring ospfd at runtime in our testbeds at Westermo. By editing its config file and restarting ospfd we see problems with route distribution. The problems started occurring in one of our testbeds when we upgraded from 0.99.17 to 0.99.20. We've tested every release since .17 and now even the latest Git, with the same result. We've managed to reduce our testbed's 14 ARM-9 routers doing a simple cost-change test, down to four Ubuntu 12.04 laptops restarting ospfd on a single laptop. Laptop setup, very simple. Two areas, one RIP network with loopback /32 nets on R1 and R4. Most of the time it works, but sometimes when we restart ospfd on R2 the edge router R1 never gets any routes. After our initial tests using OSPF and RIP we tried to reduce the topology even further by replacing RIP with another loopback /32 net. I.e., we removed R4 and replaced the 10-network with 99.99.99.99/32. Still, restarting ospfd on R2 causes R1 to sometimes not get the the 99-network. The topology for this setup is shown at the very end of this email. First, OSPF &amp; RIP setup: ______ | |---| 66.66.66.66/32 | | | R1 | | | |______| |.1 | | 172.12.0.0/24 | Area 2 | |.2 ______ | | | | | R2 | | | |______| |.1 | | 172.10.0.0/24 | Area 0 | |.2 ______ | | | ASBR | | R3 | | | |______| |.1 | | 10.1.0.0/24 | RIP |.2 ______ | |---| 14.0.0.1/32 | |---| 15.0.0.1/32 | R4 |---| 16.0.0.1/32 | | |______| In reproducing the bug we restart ospfd on R2 by pressing Ctrl-C while [ 1 ]; do sudo ./ospfd/ospfd -u root -g root -f ../ospfd.conf sleep 1; done Results ... 0.99.17: - Stable as a rock! 0.99.18 - Git: - Sometimes loss of 66-network in R3 and R4 - Loss of all RIP nets in R1 Sadly we haven't yet had the time to try and figure out where the bug is located. However, we noticed that there was quite an overhaul of the LSA refresh logic in 0.99.18. Below is some debug info: _________________ | | | R1 | |_________________| # ip route ls 169.254.0.0/16 dev eth0 scope link metric 1000 172.10.0.0/24 via 172.12.0.2 dev eth0 proto zebra metric 20 172.12.0.0/24 dev eth0 proto kernel scope link src 172.12.0.1 R1# show ip ospf route ============ OSPF network routing table ============ N IA 172.10.0.0/24 [20] area: 0.0.0.2 via 172.12.0.2, eth0 N 172.12.0.0/24 [10] area: 0.0.0.2 directly attached to eth0 ============ OSPF router routing table ============= R 172.12.0.2 [10] area: 0.0.0.2, ABR via 172.12.0.2, eth0 ============ OSPF external routing table =========== &lt;EMPTY&gt; _________________ | | | R2 | |_________________| # ip route ls 10.1.0.0/24 via 172.10.0.2 dev eth0 proto zebra metric 20 14.0.0.1 via 172.10.0.2 dev eth0 proto zebra metric 20 15.0.0.1 via 172.10.0.2 dev eth0 proto zebra metric 20 16.0.0.1 via 172.10.0.2 dev eth0 proto zebra metric 20 66.66.66.66 via 172.12.0.1 dev eth1 proto zebra metric 20 169.254.0.0/16 dev eth0 scope link metric 1000 172.10.0.0/24 dev eth0 proto kernel scope link src 172.10.0.1 172.12.0.0/24 dev eth1 proto kernel scope link src 172.12.0.2 R2# show ip ospf route ============ OSPF network routing table ============ N 172.10.0.0/24 [10] area: 0.0.0.0 directly attached to eth0 N 172.12.0.0/24 [10] area: 0.0.0.2 directly attached to eth1 ============ OSPF router routing table ============= R 10.1.0.1 [10] area: 0.0.0.0, ASBR via 172.10.0.2, eth0 R 66.66.66.66 [10] area: 0.0.0.2, ASBR via 172.12.0.1, eth1 ============ OSPF external routing table =========== N E2 10.1.0.0/24 [10/20] tag: 0 via 172.10.0.2, eth0 N E2 14.0.0.1/32 [10/20] tag: 0 via 172.10.0.2, eth0 N E2 15.0.0.1/32 [10/20] tag: 0 via 172.10.0.2, eth0 N E2 16.0.0.1/32 [10/20] tag: 0 via 172.10.0.2, eth0 N E2 66.66.66.66/32 [10/20] tag: 0 via 172.12.0.1, eth1 Second, OSPF-only setup: ______ | |---| 66.66.66.66/32 | | | R1 | | | |______| |.1 | | 172.12.0.0/24 | Area 2 | |.2 ______ | | | | | R2 | | | |______| |.1 | | 172.10.0.0/24 | Area 0 | |.2 ______ | |---| 99.99.99.99/32 | | | R3 | | | |______| Hopefully someone has the patience to read this far and has an idea about what is wrong. We are more than happy to help test patches. Regards /Joachim </pre> Joachim Nilsson 2012-05-14T15:50:12 http://comments.gmane.org/gmane.network.quagga.devel/9359 <pre>there is now a patchwork up &amp; running at http://patchwork.quagga.net/ and this is a patch to test it :) (the old patchwork at http://patchwork.diac24.net had to be flushed because the update broke the mysql data. that address is now an alias to the new one at patchwork.quagga.net. it's using postgres now.) --- FNORD | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) create mode 100644 FNORD diff --git a/FNORD b/FNORD new file mode 100644 index 0000000..824fddb --- /dev/null +++ b/FNORD &lt; at &gt;&lt; at &gt; -0,0 +1 &lt; at &gt;&lt; at &gt; +fnord! </pre> David Lamparter 2012-05-10T12:46:35 http://comments.gmane.org/gmane.network.quagga.devel/9353 <pre>Hi I am running the patch I submitted back in September 2010 since a long time now and it caused now havoc. Please consider including it into the main stream repo. </pre> Roman Hoog Antink 2012-05-09T07:35:34 http://comments.gmane.org/gmane.network.quagga.devel/9321 <pre>From: "Jorge Boncompte [DTI2]" &lt;jorge&lt; at &gt;dti2.net&gt; This patchset is part of my work in the last weeks to optimize the speed and memory usage of the bgpd daemon. My test case has been a full internet table loading on a pIII class machine. With the patches just the table loading time went from 14 to 10 seconds. And memory usage it is a lot more stable, due to less heap fragmentation. Please, take a look at it, comments welcome. Jorge Boncompte [DTI2] (25): bgpd: optimize aspath string representation and assegments handling bgpd: optimize bgp_nexthop_self() bgpd: debug buffers cleanup and optimization bgpd: remove calls to peer_sort() from fast-path bgpd: optimize loops on [e]community_hash_make() bgpd: optimize bgp_info_cmp() bgpd: remove some useless initializations bgpd: reduce attrhash_make_key() indirections bgpd: use on stack struct attr_extra on bgp_attr_aggregate_intern() bgpd: cleanup bgp_attr_unintern() bgpd: use on stack struct attr_extra in bgp_attr_unintern() bgpd: fix struct attr_extra leak in bgp_default_originate() bgpd: use on stack struct attr_extra in bgp_update_receive() bgpd: bgp_attr_default_set() already allocates a struct attr_extra bgpd: reduce struct attr_extra allocations/freeing bgpd: Remove useless initialization bgpd: optimize holdtime timer cancelling bgpd: optimize bgp_update_main() in the soft_reconfig case bgpd: fix for bgp_node_lookup() bgpd: optimize bgp_aggregate_[increment|decrement]() micro-op bgp_node_[get|lookup]() and route_node_[get|lookup]() lib: remove RUSAGE_T from struct thread lib: micro-op for thread_get() lib: do not allocate/free thread funcnames lib: optimize IPV4_ADDR_[SAME|COPY]() bgpd/bgp_aspath.c | 141 ++++++++++--------- bgpd/bgp_aspath.h | 1 + bgpd/bgp_attr.c | 121 +++++++++-------- bgpd/bgp_community.c | 19 ++- bgpd/bgp_ecommunity.c | 21 ++- bgpd/bgp_mpath.c | 2 +- bgpd/bgp_network.c | 4 +- bgpd/bgp_nexthop.c | 108 +++++++++++++--- bgpd/bgp_nexthop.h | 3 +- bgpd/bgp_packet.c | 87 +++++++----- bgpd/bgp_route.c | 357 +++++++++++++++++++++++++------------------------ bgpd/bgp_table.c | 20 ++- bgpd/bgp_vty.c | 4 +- bgpd/bgp_zebra.c | 8 +- bgpd/bgpd.c | 57 +++++---- bgpd/bgpd.h | 24 ++-- lib/memtypes.c | 2 +- lib/prefix.c | 2 +- lib/prefix.h | 4 +- lib/table.c | 18 ++- lib/thread.c | 51 +++----- lib/thread.h | 8 +- ospfd/ospf_packet.c | 2 +- 23 files changed, 594 insertions(+), 470 deletions(-) </pre> Jorge Boncompte [DTI2] 2012-05-07T17:52:50 http://comments.gmane.org/gmane.network.quagga.devel/9310 <pre>From: "Jorge Boncompte [DTI2]" &lt;jorge&lt; at &gt;dti2.net&gt; If a peer with soft-reconfiguration configured it's cleared, the function bgp_clear_route_table() doesn't free the bgp_adj_in and bgp_adj_out structures of route nodes that for some reason, ej. denied by a filter, don't have routes attached "rn-&gt;info == NULL". * bgp_route.c: (bgp_clear_route_table) moved the clearing before the allocation of the work queue items to return memory to the system allocator before allocating new one to try to avoid heap fragmentation. Signed-off-by: Jorge Boncompte [DTI2] &lt;jorge&lt; at &gt;dti2.net&gt; --- bgpd/bgp_route.c | 33 +++++++++++++++------------------ 1 files changed, 15 insertions(+), 18 deletions(-) diff --git a/bgpd/bgp_route.c b/bgpd/bgp_route.c index eae13a2..c0255b1 100644 --- a/bgpd/bgp_route.c +++ b/bgpd/bgp_route.c &lt; at &gt;&lt; at &gt; -2798,9 +2798,6 &lt; at &gt;&lt; at &gt; bgp_clear_route_table (struct peer *peer, afi_t afi, safi_t safi, struct bgp_info *ri; struct bgp_adj_in *ain; struct bgp_adj_out *aout; - - if (rn-&gt;info == NULL) - continue; /* XXX:TODO: This is suboptimal, every non-empty route_node is * queued for every clearing peer, regardless of whether it is &lt; at &gt;&lt; at &gt; -2833,6 +2830,21 &lt; at &gt;&lt; at &gt; bgp_clear_route_table (struct peer *peer, afi_t afi, safi_t safi, * this may actually be achievable. It doesn't seem to be a huge * problem at this time, */ + for (ain = rn-&gt;adj_in; ain; ain = ain-&gt;next) + if (ain-&gt;peer == peer || purpose == BGP_CLEAR_ROUTE_MY_RSCLIENT) + { + bgp_adj_in_remove (rn, ain); + bgp_unlock_node (rn); + break; + } + for (aout = rn-&gt;adj_out; aout; aout = aout-&gt;next) + if (aout-&gt;peer == peer || purpose == BGP_CLEAR_ROUTE_MY_RSCLIENT) + { + bgp_adj_out_remove (rn, aout, peer, afi, safi); + bgp_unlock_node (rn); + break; + } + for (ri = rn-&gt;info; ri; ri = ri-&gt;next) if (ri-&gt;peer == peer || purpose == BGP_CLEAR_ROUTE_MY_RSCLIENT) { &lt; at &gt;&lt; at &gt; -2848,21 +2860,6 &lt; at &gt;&lt; at &gt; bgp_clear_route_table (struct peer *peer, afi_t afi, safi_t safi, work_queue_add (peer-&gt;clear_node_queue, cnq); break; } - - for (ain = rn-&gt;adj_in; ain; ain = ain-&gt;next) - if (ain-&gt;peer == peer || purpose == BGP_CLEAR_ROUTE_MY_RSCLIENT) - { - bgp_adj_in_remove (rn, ain); - bgp_unlock_node (rn); - break; - } - for (aout = rn-&gt;adj_out; aout; aout = aout-&gt;next) - if (aout-&gt;peer == peer || purpose == BGP_CLEAR_ROUTE_MY_RSCLIENT) - { - bgp_adj_out_remove (rn, aout, peer, afi, safi); - bgp_unlock_node (rn); - break; - } } return; } </pre> Jorge Boncompte [DTI2] 2012-05-07T16:17:33 http://comments.gmane.org/gmane.network.quagga.devel/9309 <pre>From: "Jorge Boncompte [DTI2]" &lt;jorge&lt; at &gt;dti2.net&gt; bgp_afi_node_get() expects a non-NULL prd for a SAFI_MPLS_VPN prefix. * bgp_route.c: pass down the struct prefix_rd from bgp_soft_reconfig_in() and bgp_soft_reconfig_rsclient(). Signed-off-by: Jorge Boncompte [DTI2] &lt;jorge&lt; at &gt;dti2.net&gt; --- bgpd/bgp_route.c | 36 ++++++++++++++++++++++++++++-------- 1 files changed, 28 insertions(+), 8 deletions(-) diff --git a/bgpd/bgp_route.c b/bgpd/bgp_route.c index c0255b1..8dc28c8 100644 --- a/bgpd/bgp_route.c +++ b/bgpd/bgp_route.c &lt; at &gt;&lt; at &gt; -2618,7 +2618,7 &lt; at &gt;&lt; at &gt; bgp_announce_route_all (struct peer *peer) static void bgp_soft_reconfig_table_rsclient (struct peer *rsclient, afi_t afi, - safi_t safi, struct bgp_table *table) + safi_t safi, struct bgp_table *table, struct prefix_rd *prd) { struct bgp_node *rn; struct bgp_adj_in *ain; &lt; at &gt;&lt; at &gt; -2629,8 +2629,11 &lt; at &gt;&lt; at &gt; bgp_soft_reconfig_table_rsclient (struct peer *rsclient, afi_t afi, for (rn = bgp_table_top (table); rn; rn = bgp_route_next (rn)) for (ain = rn-&gt;adj_in; ain; ain = ain-&gt;next) { + struct bgp_info *ri = rn-&gt;info; + bgp_update_rsclient (rsclient, afi, safi, ain-&gt;attr, ain-&gt;peer, - &amp;rn-&gt;p, ZEBRA_ROUTE_BGP, BGP_ROUTE_NORMAL, NULL, NULL); + &amp;rn-&gt;p, ZEBRA_ROUTE_BGP, BGP_ROUTE_NORMAL, prd, + (bgp_info_extra_get (ri))-&gt;tag); } } &lt; at &gt;&lt; at &gt; -2641,18 +2644,25 &lt; at &gt;&lt; at &gt; bgp_soft_reconfig_rsclient (struct peer *rsclient, afi_t afi, safi_t safi) struct bgp_node *rn; if (safi != SAFI_MPLS_VPN) - bgp_soft_reconfig_table_rsclient (rsclient, afi, safi, NULL); + bgp_soft_reconfig_table_rsclient (rsclient, afi, safi, NULL, NULL); else for (rn = bgp_table_top (rsclient-&gt;bgp-&gt;rib[afi][safi]); rn; rn = bgp_route_next (rn)) if ((table = rn-&gt;info) != NULL) - bgp_soft_reconfig_table_rsclient (rsclient, afi, safi, table); + { + struct prefix_rd prd; + prd.family = AF_UNSPEC; + prd.prefixlen = 64; + memcpy(&amp;prd.val, rn-&gt;p.u.val, 8); + + bgp_soft_reconfig_table_rsclient (rsclient, afi, safi, table, &amp;prd); + } } static void bgp_soft_reconfig_table (struct peer *peer, afi_t afi, safi_t safi, - struct bgp_table *table) + struct bgp_table *table, struct prefix_rd *prd) { int ret; struct bgp_node *rn; &lt; at &gt;&lt; at &gt; -2666,9 +2676,12 &lt; at &gt;&lt; at &gt; bgp_soft_reconfig_table (struct peer *peer, afi_t afi, safi_t safi, { if (ain-&gt;peer == peer) { + struct bgp_info *ri = rn-&gt;info; + ret = bgp_update (peer, &amp;rn-&gt;p, ain-&gt;attr, afi, safi, ZEBRA_ROUTE_BGP, BGP_ROUTE_NORMAL, - NULL, NULL, 1); + prd, (bgp_info_extra_get (ri))-&gt;tag, 1); + if (ret &lt; 0) { bgp_unlock_node (rn); &lt; at &gt;&lt; at &gt; -2689,12 +2702,19 &lt; at &gt;&lt; at &gt; bgp_soft_reconfig_in (struct peer *peer, afi_t afi, safi_t safi) return; if (safi != SAFI_MPLS_VPN) - bgp_soft_reconfig_table (peer, afi, safi, NULL); + bgp_soft_reconfig_table (peer, afi, safi, NULL, NULL); else for (rn = bgp_table_top (peer-&gt;bgp-&gt;rib[afi][safi]); rn; rn = bgp_route_next (rn)) if ((table = rn-&gt;info) != NULL) -bgp_soft_reconfig_table (peer, afi, safi, table); + { + struct prefix_rd prd; + prd.family = AF_UNSPEC; + prd.prefixlen = 64; + memcpy(&amp;prd.val, rn-&gt;p.u.val, 8); + + bgp_soft_reconfig_table (peer, afi, safi, table, &amp;prd); + } } </pre> Jorge Boncompte [DTI2] 2012-05-07T16:17:34 http://comments.gmane.org/gmane.network.quagga.devel/9308 <pre>From: "Jorge Boncompte [DTI2]" &lt;jorge&lt; at &gt;dti2.net&gt; ospfd was crashing some times on neighbour going down. The cause was that ospf_nsm_event() was accessing already freed memory in ospf_nbr_delete() call from ospf_nsm_event(). What happens is that since commit b5043aab (lib: fix incorrect thread list...) now a thread can be on the event and ready lists but thread_cancel_event() doesn't account for that. * thread.c: (thread_cancel_event) loop on the ready list too to cancel pending events. Signed-off-by: Jorge Boncompte [DTI2] &lt;jorge&lt; at &gt;dti2.net&gt; --- lib/thread.c | 18 ++++++++++++++++++ 1 files changed, 18 insertions(+), 0 deletions(-) diff --git a/lib/thread.c b/lib/thread.c index b36c43a..dd0413b 100644 --- a/lib/thread.c +++ b/lib/thread.c &lt; at &gt;&lt; at &gt; -916,6 +916,24 &lt; at &gt;&lt; at &gt; thread_cancel_event (struct thread_master *m, void *arg) thread_add_unuse (m, t); } } + + /* thread can be on the ready list too */ + thread = m-&gt;ready.head; + while (thread) + { + struct thread *t; + + t = thread; + thread = t-&gt;next; + + if (t-&gt;arg == arg) + { + ret++; + thread_list_delete (&amp;m-&gt;ready, t); + t-&gt;type = THREAD_UNUSED; + thread_add_unuse (m, t); + } + } return ret; } </pre> Jorge Boncompte [DTI2] 2012-05-07T16:17:31 http://comments.gmane.org/gmane.network.quagga.devel/9307 <pre>From: "Jorge Boncompte [DTI2]" &lt;jorge&lt; at &gt;dti2.net&gt; * bgp_route.c: (bgp_announce_check) check that the peer address it is not the same as the nexthop address per RFC4271 5.1.3. Fix the IPV6 check to use the rigth addresses. Signed-off-by: Jorge Boncompte [DTI2] &lt;jorge&lt; at &gt;dti2.net&gt; --- bgpd/bgp_route.c | 10 ++++++---- 1 files changed, 6 insertions(+), 4 deletions(-) diff --git a/bgpd/bgp_route.c b/bgpd/bgp_route.c index d96224d..eae13a2 100644 --- a/bgpd/bgp_route.c +++ b/bgpd/bgp_route.c &lt; at &gt;&lt; at &gt; -803,13 +803,15 &lt; at &gt;&lt; at &gt; bgp_announce_check (struct bgp_info *ri, struct peer *peer, struct prefix *p, if (from == peer) return 0; - /* If peer's id and route's nexthop are same. draft-ietf-idr-bgp4-23 5.1.3 */ + /* If peer's id or address and route's nexthop are the same. RFC4271 5.1.3 */ if (p-&gt;family == AF_INET - &amp;&amp; IPV4_ADDR_SAME(&amp;peer-&gt;remote_id, &amp;riattr-&gt;nexthop)) + &amp;&amp; (IPV4_ADDR_SAME(&amp;peer-&gt;remote_id, &amp;riattr-&gt;nexthop) || + IPV4_ADDR_SAME(&amp;peer-&gt;su_remote-&gt;sin.sin_addr, &amp;riattr-&gt;nexthop))) return 0; #ifdef HAVE_IPV6 - if (p-&gt;family == AF_INET6 - &amp;&amp; IPV6_ADDR_SAME(&amp;peer-&gt;remote_id, &amp;riattr-&gt;nexthop)) + if (p-&gt;family == AF_INET6 &amp;&amp; riattr-&gt;extra + &amp;&amp; IPV6_ADDR_SAME(&amp;peer-&gt;su_remote-&gt;sin6.sin6_addr, +&amp;riattr-&gt;extra-&gt;mp_nexthop_global)) return 0; #endif </pre> Jorge Boncompte [DTI2] 2012-05-07T16:17:32 http://comments.gmane.org/gmane.network.quagga.devel/9305 <pre>Minor build fixes for building without IPv6/RTADV support. Regards /Joachim </pre> Joachim Nilsson 2012-05-07T09:30:15 http://comments.gmane.org/gmane.network.quagga.devel/9303 <pre>Hello, I am a student working on some extensions for ospfv3 autoconfig and I need to be able to change the router-id at any moment in case of duplicate router-id's on the network I can detect when there are duplicates on the network and decide which router changes but I can't make the change correctly. What I do is change the router-id fields in the ospf6 structure: ... generate new router id ... ospf6-&gt;router_id_static = new_router_id; ospf6-&gt;router_id = new_router_id; And this seems to work because the other routers correctly update their spf tree but something goes wrong in the routes in the changing router. So what should be done after this to ensure that everything else takes the change into account? something in zebra daemon? I suppose some things should be purged and rebuilt but I can't seem to find what, I tried some things like: ospf6_lsdb_remove_all (o-&gt;lsdb); or ospf6_spf_init (); But nothing seems to work. And also when I try to change through the console: lubuntu1OSPF6(config-ospf6)# router-id 0.0.0.142 Has no effet at all beside changing ospf6-&gt;router_id_static only. By the way, what is the difference between router_id and router_id_static? David Dyja d.dyja &lt; at &gt;student.ulg.ac.be </pre> David Dyja 2012-05-06T18:55:32 http://comments.gmane.org/gmane.network.quagga.devel/9296 <pre>For my many sins I have spent the day looking at the SAFI_MPLS_VPN stuff. I note that when constructing an MP_REACH and an MP_UNREACH that the tag part does not appear to have the "Bottom of Stack" bit set. Now, this may be because somewhere in the RFCs it says that there will only ever be one label in the stack for MP_REACH. Of course, MP_UNREACH goes with a single NULL Label in any case... Can anyone point me in the direction of the RFC which says that "Bottom of Stack" is not required ? Or have I missed the piece of code which sets it ? Thanks, Chris </pre> 'Chris Hall' 2012-05-05T18:53:32 http://comments.gmane.org/gmane.network.quagga.devel/9293 <pre>Got an odd issue with IPv6 ospf6d. I have a couple quagga routers (0.99.20.1 on linux) that share a network with a Cisco router that is functioning as Area Border Router. The Area is 131.94.128.0 The cisco Area Border Router also connects to area 0. In area 0 is a ASBR that is injecting a default route (::0/0) When I bring up router1 all seems fine. Router1 receives and uses a few routes from area 0, and announces a downstream network as well. But as soon as I bring up router 2, after the DB exchange a LSA is sent from router2 to router1, and this causes router1 to do a SPF calculation. When router1 does this SPF calculation it loses the default route. Well, it discards the default route. router2 comes up fine, and it does use the ASE default route. Until something changes and a SPF calc is done on router2, then it will discard the default route. Any ideas what is happening here? It doesn't make sense to me other than some bug in the SPF calculation. But maybe I am missing something?? Details available via web in directory http://users.cis.fiu.edu/~esj/quagga-problem-ospf3/ http://users.cis.fiu.edu/~esj/quagga-problem-ospf3/network.png is the picture http://users.cis.fiu.edu/~esj/quagga-problem-ospf3/router1.ospf6d.conf http://users.cis.fiu.edu/~esj/quagga-problem-ospf3/router2.ospf6d.conf are the relevant parts of the config files of the two quagga routers E </pre> Eric S. Johnson 2012-05-04T20:51:53 http://comments.gmane.org/gmane.network.quagga.devel/9281 <pre>Hi everyone! Anyone planning on adding support for setting metric on static routes in Zebra? Or is this redundant, considering the possibility to set a distance instead? Also, I read earlier that there were some thoughts on merging Static and Kernel routes in some way or form. Or did I dream this? Regards /Joachim </pre> Joachim Nilsson 2012-05-04T09:56:31 Search the mailing list at Gmane query http://search.gmane.org/?group=$group=gmane.network.quagga.devel