- Bird-users - Mailing Lists

Bloom-bird: a scalable open-source router based on Bloom Filter
by Bahram BahramBeigy 10 Jan '15

10 Jan '15

We have accelerated FIB lookups (fib_find() and fib_route()) using a data structure named Bloom Filter when number of inserted nodes into FIB becomes huge, for example more than 200,000 IPs are inserted into one FIB. Consequently, the linked list chains becomes huge, the Bloom filter avoids traverse these long chains when an IP cannot be found. The paper: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6999678 Please let me know your opinion about it. Or any suggestions ... Thanks in advance. -- Best wishes, Bahram BahramBeigy http://www.bahrambeigy.ir

3 2

Fwd: Bloom-bird: a scalable open-source router based on Bloom Filter
by Bahram BahramBeigy 09 Jan '15

09 Jan '15

The paper is available through our Researchgate page: https://www.researchgate.net/publication/267514772_Bloom-Bird_A_Scalable_Op… On Fri, Jan 9, 2015 at 6:17 PM, Sargun Dhillon <sargun(a)sargun.me> wrote: > I don't have an IEEE membership. Do you have a link to a non pay-walled > PDF? > > On Fri, Jan 9, 2015 at 10:01 AM, Bahram BahramBeigy <bahramwhh(a)gmail.com> > wrote: > > We have accelerated FIB lookups (fib_find() and fib_route()) using a data > > structure named Bloom Filter when number of inserted nodes into FIB > becomes > > huge, for example more than 200,000 IPs are inserted into one FIB. > > Consequently, the linked list chains becomes huge, the Bloom filter > avoids > > traverse these long chains when an IP cannot be found. > > > > The paper: > > http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6999678 > > > > Please let me know your opinion about it. > > Or any suggestions ... > > > > Thanks in advance. > > > > -- > > Best wishes, > > Bahram BahramBeigy > > http://www.bahrambeigy.ir > > > -- Best wishes, Bahram BahramBeigy http://www.bahrambeigy.ir -- Best wishes, Bahram BahramBeigy http://www.bahrambeigy.ir

1 0

Bloom-bird: a scalable open-source router based on Bloom Filter
by Bahram BahramBeigy 09 Jan '15

09 Jan '15

We have accelerated FIB lookups (fib_find() and fib_route()) using a data structure named Bloom Filter when number of inserted nodes into FIB becomes huge, for example more than 200,000 IPs are inserted into one FIB. Consequently, the linked list chains becomes huge, the Bloom filter avoids traverse these long chains when an IP cannot be found. The paper: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6999678 Please let me know your opinion about it. Or any suggestions ... Thanks in advance. -- Best wishes, Bahram BahramBeigy http://www.bahrambeigy.ir

1 0

QNX port of Bird 1.4.5, corrupt linked lists?
by Pendzik, Edward 11 Dec '14

11 Dec '14

Thu Dec 11 01:46:41 EST 2014 Hello. I am working on porting bird 1.4.5 to QNX SP1 (QNX OS service pack 1). QNX is actually close to BSD and there have been very few changes to the code. I have bird built but I get a core dump on initialization. I was hoping that someone could give me some specific ideas on exactly where to look while debugging. I have many years of writing and debugging Linux and Solaris drivers and networking code, but I am not yet familiar with the bird internal data structures. I tried building with and without pthreads. I get a core dump both ways, in different spots, both when calling WALK_LIST() with a list. // with pthreads #0 0x00111460 in if_notify_change (c=1, i=0x187050) at ../../nest/iface.c:218 218 WALK_LIST(p, active_proto_list) // without pthreads if_start_update () at ../../nest/iface.c:338 338 WALK_LIST(i, iface_list) I also turned off the '#define POISON' in lib/slab.c, and on a different run did a bzero() instead of a memset(POISON). Neither changed the results. Attached find my bird.conf Any hints would be appreciated. PS Something else weird. I get an error when bird starts up and parses bird.conf. 'protocol device' rejects 'interface "en0";', even though en0 is the correct name of my ethernet interface, and ifconfig shows it exists and is up with an ipaddr. protocol device { interface "en0"; } % ifconfig lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 33192 inet 127.0.0.1 netmask 0xff000000 wip0: flags=8802<BROADCAST,SIMPLEX,MULTICAST> mtu 1500 address: 01 en0: flags=80008843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST,SHIM> mtu 1500 address: 00:80:0f:9e:08:74 media: Ethernet 100baseTX full-duplex status: active inet 192.168.1.1 netmask 0xffffff00 broadcast 192.168.1.255 Thanks, Ed Pendzik ependzik(a)harris.com //------------------------------------------------------------ btop="/fs" confopts="\ --sysconfdir=$btop \ --localstatedir=$btop \ --prefix=$btop \ CC=$HOME/qccbird \ CFLAGS=-V4.4.2,gcc_ntoarmle \ --disable-pthreads \ --build=i686-pc-linux-gnu \ --host=arm-unknown-nto-qnx6.5.0 \ --with-sysconfig=qnx \ --with-protocols=ospf,pipe,static,bfd,rip \ --enable-debug \ --disable-client \ " ./configure $confopts //------------------------------------------------------------ // WITH PTHREADS (--enable-pthreads) // console output 117.localhost_root 2014-12-11 06:10:06 <TRACE> direct1: Initializing 2014-12-11 06:10:06 <TRACE> kernel1: Initializing 2014-12-11 06:10:06 <TRACE> device1: Initializing 2014-12-11 06:10:06 <TRACE> ospf1: Initializing 2014-12-11 06:10:06 <TRACE> device1: Starting 2014-12-11 06:10:06 <TRACE> device1: Scanning interfaces @@@ if_notify_change: this is debug() @@@ if_notify_change: this is DBG() Interface change notification (1) for lo0 IF1: lo0 UP LINK-UP MA LOOP IGN MTU=33192 127.0.0.1, net 127.0.0.0/8 bc 0.0.0.0 -> 0.0.0.0 DOWN unlinking old /fs/coredumps/bird.core dumping to /fs/coredumps/bird.core thread 1.flags is 0x1 thread 1 is was SIGNALLED thread 1.why is 0x1 ldd mapinfos: Thread 1's SP is at 0xff7d4 Adjusting 24576 bytes @ 0xfa000->0x100000 to 4096 bytes @ 0xff000->0x100000 thread 1 is current thread! blanking 24576 bytes of stack memory at 0xff000 max=0x6000, min = 0x5000 (0xff000->0x100000) dumping 483328 bytes of memory at 0x100000 dumping 8192 bytes of memory at 0x176000 dumping 4096 bytes of memory at 0x178000 dumping 135168 bytes of memory at 0x179000 dumping 471040 bytes of memory at 0x1000000 dumping 16384 bytes of memory at 0x1073000 dumping 4096 bytes of memory at 0x1077000 dumping 172032 bytes of memory at 0x78000000 dumping 24576 bytes of memory at 0x7802a000 ntoarm-gdb bird bird.core GNU gdb 6.8 qnx-nto (rev. 506) Copyright (C) 2008 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "--host=i686-pc-linux-gnu --target=arm-unknown-nto-qnx6.5.0"... Reading symbols from /home2/opt/qnx650sp1/target/qnx6/armle/lib/libsocket.so.3...done. Loaded symbols for /home2/opt/qnx650sp1/target/qnx6/armle/lib/libsocket.so.3 Reading symbols from /home2/opt/qnx650sp1/target/qnx6/armle/lib/libc.so.3...done. Loaded symbols for /home2/opt/qnx650sp1/target/qnx6/armle/lib/libc.so.3 Program terminated with signal 10, Bus error. [New pid 1708126 tid 1] #0 0x00111460 in if_notify_change (c=1, i=0x187050) at ../../nest/iface.c:218 218 WALK_LIST(p, active_proto_list) (gdb) whete Undefined command: "whete". Try "help". (gdb) where #0 0x00111460 in if_notify_change (c=1, i=0x187050) at ../../nest/iface.c:218 #1 0x00111624 in if_change_flags (i=0x187050, flags=1073742306) at ../../nest/iface.c:254 #2 0x00111988 in if_end_partial_update (i=0x187050) at ../../nest/iface.c:350 #3 0x00111a34 in if_end_update () at ../../nest/iface.c:368 #4 0x0015a184 in kif_do_scan (p=0x18a528) at krt-sock.c:894 #5 0x0015ab0c in kif_scan (t=0x1860d0) at krt.c:98 #6 0x0015af64 in kif_start (P=0x18a528) at krt.c:191 #7 0x0010e22c in proto_rethink_goal (p=0x18a528) at ../../nest/proto.c:632 #8 0x0010dff0 in protos_commit (new=0x18f9a8, old=0x0, force_reconfig=0, type=1) at ../../nest/proto.c:580 #9 0x00151df0 in config_do_commit (c=0x18f9a8, type=1) at conf.c:255 #10 0x00152030 in config_commit (c=0x18f9a8, type=1, timeout=0) at conf.c:348 #11 0x0015f788 in main (argc=2, argv=0xff944) at main.c:822 (gdb) //------------------------------------------------------------ // WITHOUT PTHREADS (--disable-pthreads) // console output 128.localhost_root 2014-12-11 06:15:41 <TRACE> direct1: Initializing 2014-12-11 06:15:41 <TRACE> kernel1: Initializing 2014-12-11 06:15:41 <TRACE> device1: Initializing 2014-12-11 06:15:41 <TRACE> ospf1: Initializing 2014-12-11 06:15:41 <TRACE> device1: Starting 2014-12-11 06:15:41 <TRACE> device1: Scanning interfaces dumping to /fs/coredumps/bird.core thread 1.flags is 0x1 thread 1 is was SIGNALLED thread 1.why is 0x1 ldd mapinfos: Thread 1's SP is at 0xff828 Adjusting 24576 bytes @ 0xfa000->0x100000 to 4096 bytes @ 0xff000->0x100000 thread 1 is current thread! blanking 24576 bytes of stack memory at 0xff000 max=0x6000, min = 0x5000 (0xff000->0x100000) dumping 483328 bytes of memory at 0x100000 dumping 8192 bytes of memory at 0x176000 dumping 4096 bytes of memory at 0x178000 dumping 98304 bytes of memory at 0x179000 dumping 471040 bytes of memory at 0x1000000 dumping 16384 bytes of memory at 0x1073000 dumping 4096 bytes of memory at 0x1077000 dumping 172032 bytes of memory at 0x78000000 dumping 24576 bytes of memory at 0x7802a000 ntoarm-gdb bird bird.core GNU gdb 6.8 qnx-nto (rev. 506) Copyright (C) 2008 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "--host=i686-pc-linux-gnu --target=arm-unknown-nto-qnx6.5.0"... warning: Host file /home2/opt/qnx650sp1/target/qnx6/armle/lib/libsocket.so.3 does not match target file. warning: Host file /home2/opt/qnx650sp1/target/qnx6/armle/lib/libc.so.3 does not match target file. Reading symbols from /home2/opt/qnx650sp1/target/qnx6/armle/lib/libsocket.so.3...done. Loaded symbols for /home2/opt/qnx650sp1/target/qnx6/armle/lib/libsocket.so.3 Reading symbols from /home2/opt/qnx650sp1/target/qnx6/armle/lib/libc.so.3...done. Loaded symbols for /home2/opt/qnx650sp1/target/qnx6/armle/lib/libc.so.3 Program terminated with signal 11, Segmentation fault. [New pid 2158686 tid 1] #0 0x001118c0 in if_start_update () at ../../nest/iface.c:338 338 WALK_LIST(i, iface_list) (gdb) bt #0 0x001118c0 in if_start_update () at ../../nest/iface.c:338 #1 0x0015a0fc in kif_do_scan (p=0x18a528) at krt-sock.c:892 #2 0x0015aa9c in kif_scan (t=0x1860d0) at krt.c:98 #3 0x0015aef4 in kif_start (P=0x18a528) at krt.c:191 #4 0x0010e1bc in proto_rethink_goal (p=0x18a528) at ../../nest/proto.c:632 #5 0x0010df80 in protos_commit (new=0x18f9a8, old=0x0, force_reconfig=0, type=1) at ../../nest/proto.c:580 #6 0x00151d80 in config_do_commit (c=0x18f9a8, type=1) at conf.c:255 #7 0x00151fc0 in config_commit (c=0x18f9a8, type=1, timeout=0) at conf.c:348 #8 0x0015f6cc in main (argc=2, argv=0xff944) at main.c:822 (gdb)

2 1

'Netlink: File exists' messages out of bird6
by Michael Fincham 01 Dec '14

01 Dec '14

Hi all, I'm periodically seeing bird6 log these message (and get rate limited, which implies there were quite a few at once...) Nov 30 03:38:17 wlg-por-router2 bird6: Netlink: File exists Nov 30 03:38:17 wlg-por-router2 bird6: ... Any hints on how best to troubleshoot this? -- Michael

2 1

RTD_BLACKHOLE overridden by matching route.
by Thomas Johnson 29 Nov '14

29 Nov '14

I am having some difficulty blackholing a route. The route in question is received from a multihop BGP session (emulating cymru), and I am setting the dest attribute to RTD_BLACKHOLE. However, if the same route is received via another BGP session, the blackhole is overriden, as shown here: bird> show route 10.0.2.0/24 all 10.0.2.0/24 via 38.8.8.105 on em1 [bgp_cogent 17:38:53] * (100) [AS174i] Type: BGP unicast univ BGP.origin: IGP BGP.as_path: 174 BGP.next_hop: 38.8.8.105 BGP.local_pref: 100 blackhole [l3_bogons 18:00:03 from 10.0.1.255] (100) [AS3356i] Type: BGP unicast univ BGP.origin: IGP BGP.as_path: 3356 BGP.next_hop: 10.0.1.255 BGP.local_pref: 255 BGP.community: (65332,888) The relevant parts of my test configuration are below. Is there some trick to making the blackhole "absolute"? filter bgp_in_cogent { if ! rt_import_all(174) then reject; accept; } template bgp cymru_bogons_full { local as 68; multihop 255; export none; import filter { bgp_community.add((65332,888)); dest = RTD_BLACKHOLE; bgp_local_pref = 255; accept; }; } protocol bgp bgp_cogent from ebgp { neighbor 38.8.8.105 as 174; import filter bgp_in_cogent; } protocol bgp l3_bogons from cymru_bogons_full { local 162.9.5.3 as 68; neighbor 10.0.1.255 as 3356; } Thank you!

2 1

BFD to Juniper SRX
by Tom Harbert 27 Nov '14

27 Nov '14

Hello, I am running BIRD 1.4.0 on an Ubuntu linux machine which peers BGP successfully with a Juniper SRX firewall. I am having difficulties getting BFD past the Init state. See below for both BIRD and SRX config and logs. It appears as though the firewall is not receiving the BFD messages. I have confirmed that there are no access control restrictions for BFD hitting the SRX (host-inbound-traffic as well as loopback filter). I have already adjusted the Linux source port selection with: sysctl -w net.ipv4.ip_local_port_range="49152 65535" I am sure its something basic, troubleshooting suggestions appreciated. Thanks, Tom. // bird.conf protocol bfd { interface "172.30.6.8/32" { interval 1000 ms; multiplier 3; }; } protocol bgp iBGP_1 { .... neighbor 172.30.6.1 as z; source address 172.30.6.8; .... bfd; .... } // bird host $ sudo ip addr show eth0 | grep inet inet 172.30.6.8/24 brd 172.30.6.255 scope global eth0 $ sudo birdc show bfd session BIRD 1.4.0 ready. bfd1: IP address Interface State Since Interval Timeout 172.30.6.1 --- Init 05:27:46 1.000 3.000 $ sudo tcpdump -i eth0 udp port 3784 tcpdump: verbose output suppressed, use -v or -vv for full protocol decode listening on eth0, link-type EN10MB (Ethernet), capture size 65535 bytes 05:47:35.873663 IP srx-host.49152 > bird-host.3784: BFDv1, Control, State Down, Flags: [none], length: 24 05:47:36.642025 IP srx-host.49152 > bird-host.3784: BFDv1, Control, State Down, Flags: [none], length: 24 05:47:37.521546 IP srx-host.49152 > bird-host.3784: BFDv1, Control, State Down, Flags: [none], length: 24 05:47:38.349886 IP srx-host.49152 > bird-host.3784: BFDv1, Control, State Down, Flags: [none], length: 24 05:53:19.824960 IP (tos 0xc0, ttl 255, id 16504, offset 0, flags [none], proto UDP (17), length 52) srx-host.49152 > bird-host.3784: BFDv1, length: 24 Control, State Down, Flags: [none], Diagnostic: No Diagnostic (0x00) Detection Timer Multiplier: 3 (3000 ms Detection time), BFD Length: 24 My Discriminator: 0x0000000f, Your Discriminator: 0x00000000 Desired min Tx Interval: 1000 ms Required min Rx Interval: 1000 ms Required min Echo Interval: 0 ms $ sudo tail -f /var/log/bird.log | grep bfd 2014-11-27 05:27:42 <TRACE> bfd1: Sending CTL to 172.30.6.1 [Down] 2014-11-27 05:27:43 <TRACE> bfd1: Sending CTL to 172.30.6.1 [Down] 2014-11-27 05:27:43 <TRACE> bfd1: CTL received from 172.30.6.1 [AdminDown] 2014-11-27 05:27:43 <TRACE> bfd1: Sending CTL to 172.30.6.1 [Down] 2014-11-27 05:27:43 <TRACE> bfd1: CTL received from 172.30.6.1 [AdminDown] 2014-11-27 05:27:43 <TRACE> bfd1: Sending CTL to 172.30.6.1 [Down] 2014-11-27 05:27:43 <TRACE> bfd1: CTL received from 172.30.6.1 [AdminDown] 2014-11-27 05:27:43 <TRACE> bfd1: Sending CTL to 172.30.6.1 [Down] 2014-11-27 05:27:43 <TRACE> bfd1: CTL received from 172.30.6.1 [AdminDown] 2014-11-27 05:27:46 <TRACE> bfd1: CTL received from 172.30.6.1 [Down] 2014-11-27 05:27:46 <TRACE> bfd1: Session to 172.30.6.1 changed state from Down to Init 2014-11-27 05:27:46 <TRACE> bfd1: Sending CTL to 172.30.6.1 [Init] 2014-11-27 05:27:46 <TRACE> bfd1: CTL received from 172.30.6.1 [Down] 2014-11-27 05:27:46 <TRACE> bfd1: Sending CTL to 172.30.6.1 [Init] 2014-11-27 05:27:46 <TRACE> bfd1: Sending CTL to 172.30.6.1 [Init] 2014-11-27 05:27:46 <TRACE> bfd1: CTL received from 172.30.6.1 [Down] // juniper firewall set protocols bgp group X bfd-liveness-detection minimum-interval 1000 set security zones security-zone X interfaces reth2.106 host-inbound-traffic protocols bfd > show bfd session Detect Transmit Address State Interface Time Interval Multiplier 172.30.6.8 Down reth2.106 0.000 1.000 3 172.30.6.9 Down reth2.106 0.000 1.000 3 2 sessions, 2 clients Cumulative transmit rate 2.0 pps, cumulative receive rate 0.0 pps // traceoptions Nov 27 16:25:23.540758 Initiated BFD session to peer 172.30.6.8 (Internal AS 65000): address=172.30.6.8 ifindex=151 ifname=reth2.106 txivl=1000 rxivl=1000 mult=3 ver=255 Nov 27 16:25:27.562265 Initiated BFD session to peer 172.30.6.9 (Internal AS 65000): address=172.30.6.9 ifindex=151 ifname=reth2.106 txivl=1000 rxivl=1000 mult=3 ver=255 Nov 27 16:26:59.795599 Terminated BFD session to peer 172.30.6.8 (Internal AS 65000) (Closing) Nov 27 16:26:59.798422 Terminated BFD session to peer 172.30.6.9 (Internal AS 65000) (Closing) Nov 27 16:27:31.810248 Initiated BFD session to peer 172.30.6.8 (Internal AS 65000): address=172.30.6.8 ifindex=151 ifname=reth2.106 txivl=1000 rxivl=1000 mult=3 ver=255 Nov 27 16:27:35.811656 Initiated BFD session to peer 172.30.6.9 (Internal AS 65000): address=172.30.6.9 ifindex=151 ifname=reth2.106 txivl=1000 rxivl=1000 mult=3 ver=255

2 1

Source address selection for transit-router with "blackholed" linknet
by oskar＠cetex.se 13 Nov '14

13 Nov '14

Hi! I'm working on getting a few linux-routers with bird up and running and I'm wondering: Is there any way to influence source-address selection for the routes installed by bird? (how do other bird-users handle this?, I guess I'm not alone?) We have a transit from a provider that doesn't announce the linknet, as a lot of providers do. (and i shouldn't source traffic from the linknet, i should use own addresses) When i source traffic from the server/router it seems linux default-action is to select the interface closest to the destination which is the peering-linknet and it happily sends tcp / udp / icmp requests to the world. The replies never make it back though since the ip isn't reachable from anywhere else. With ip route there seems to be this flag "src <addr>", so you can specify "ip route add x.x.x.x/y via z.z.z.z src a.a.a.a", this influences the kernel to select src a.a.a.a when sending traffic to x.x.x.x/y, but i haven't found such an option in the kernel-table in bird. If i have a router where only the "inside"-ip is reachable from the internet, and all routes point to the outside, how do i make it source all locally generated traffic from the inside-ip? Do you setup multiple routing-tables and set the default-table to direct traffic to the inside, then on the inside-interface have the full routing table (would this work?), or how do you handle this? Best regards Oskar Stenman

3 2

Bird on ubuntu
by Prashant Chougule 13 Nov '14

13 Nov '14

Hi, I have two ubuntu(14.04) hosts running bird(1.4.0) on it, my bird.conf is as follows, *On Host-1*, router id 192.168.1.101; protocol device { scan time 10; } protocol kernel { export all; scan time 15; } protocol static static_bgp { import all; } protocol bgp { import all; export where proto = "static_bgp"; local as 200; neighbor 10.0.6.2 as 1000; } *On Host-2* router id 192.168.1.102; protocol device { scan time 10; } protocol kernel { export all; scan time 15; } protocol static static_bgp { import all; } protocol bgp { import all; export where proto = "static_bgp"; local as 1000; neighbor 10.0.6.1 as 200; } I am not able establish connection between them,it say , no route to host in last error message. how do I configure bird between two ubuntu hosts? -- Thanks &Regards, Prashant Chougule.

2 1

How to export address from loopback via BGP
by Eugene Prokopiev 11 Nov '14

11 Nov '14

Hi, How to export address 1.1.1.1/32 from lo interface via BGP? I tried to do: router id 2.2.2.2; protocol kernel { } protocol static { route 1.1.1.1/32 via "lo"; } protocol bgp { local as 65000; neighbor 2.2.2.1 as 65000; export filter { accept; }; } but there are no any prefixes on neighbor 2.2.2.1 :( -- WBR, Eugene Prokopiev

5 7