- Bird-users - Mailing Lists

Re: problems with simple configuration
by David Rohleder 09 Apr '01

09 Apr '01

Ondrej Filip <feela(a)ipex.cz> writes: > On 9 Apr 2001, David Rohleder wrote: > > > > > Hello, > > > > i have got a simple config, which behaves badly: > > > > Internet P-t-P netw. local network > > ------------| VNR |----------| R |------------- > > > > > > ---[/etc/bird.conf > > > > debug protocols all; > > > > protocol direct { > > # interface "-eth*", "*"; # Restrict network interfaces it works with > > } > > > > protocol kernel { > > # learn; # Learn all alien routes from the kernel > > persist; # Don't remove routes on bird shutdown > > scan time 20; # Scan kernel routing table every 20 seconds > > # import none; # Default is import all > > export all; # Default is export none > > # kernel table 5; # Kernel table to synchronize with (default: mai > > n) > > } > > > > protocol static { > > } > > > > protocol rip MyRIP { # You can also use an explicit name > > port 520; > > interface "eth2" { mode broadcast; }; > > authentication none; > > import none; > > export all; > > } > > > > protocol ospf MyOSPF { > > # rfc1583compat yes; > > area 0.0.0.2 { > > # stub no; > > # tick 1; > > interface "eth0" { > > cost 500; > > dead count 4; > > type broadcast; > > }; > > interface "eth1" { > > cost 500; > > dead count 4; > > type broadcast; > > }; > > interface "eth2" { > > stub yes; > > cost 10; > > }; > > interface "eth3" { > > stub yes; > > cost 10; > > }; > > }; > > } > > > > ------[/etc/bird.conf > > > > eth1 and eth3 are up, but there are not any machines. eth3 doesn't > > have IP address. > > > > > > After start, VNR (Bay Networks) doesn't see internal network. But they > > are in full state! > > > > And even more strange thing: VNR learns LSA after some action in > > birdc. For example, when i run debug MyOSPF all, or even just start birdc! > > > > May be some timing problems. It is PIII/600, 128MB RAM. > > > > some strange messages: > > > > MyOSPF: LS ack sent to 147.251.61.1 (delayed) > > > > shows every max. 20 sec. > > Hmm it's very strange. :-( Please could you tell me what says 'show ospf' > and other 'show ospf *'? After power-on: BIRD 1.0.4 ready. bird> show ospf MyOSPF: Number of areas: 1 Area: 0.0.0.2 (2) Stub: No RT scheduler tick: 7 Number of interfaces: 3 Number of LSAs in DB: 308 Number of neighbors: 1 Number of adjacent neighbors: 1 bird> show ospf nei MyOSPF: Router ID Pri State DTime Router IP Interface 147.251.240.166 1 full/dr 00:38 147.251.61.1 eth0 bird> show ospf int MyOSPF: Interface "eth0": Area: 0.0.0.2 (2) Type: broadcast State: backup Priority: 1 Cost: 500 Hello timer: 10 Wait timer: 40 Dead timer: 40 Retransmit timer: 5 Designed router (ID): 147.251.240.166 Designed router (IP): 147.251.61.1 Backup designed router (ID): 147.251.5.1 Backup designed router (IP): 147.251.61.2 Interface "eth1": Area: 0.0.0.2 (2) Type: broadcast State: dr Priority: 1 Cost: 500 Hello timer: 10 Wait timer: 40 Dead timer: 40 Retransmit timer: 5 Designed router (ID): 147.251.5.1 Designed router (IP): 147.251.61.6 Backup designed router (ID): 0.0.0.0 Backup designed router (IP): 0.0.0.0 Interface "eth2": Area: 0.0.0.2 (2) Type: broadcast State: dr (stub) Priority: 1 Cost: 10 Hello timer: 10 Wait timer: 40 Dead timer: 40 Retransmit timer: 5 Designed router (ID): 147.251.5.1 Designed router (IP): 147.251.5.1 Backup designed router (ID): 0.0.0.0 Backup designed router (IP): 0.0.0.0 After some time (~1 min, without any other actions), bird begins to work: bird> show ospf MyOSPF: Number of areas: 1 Area: 0.0.0.2 (2) Stub: No RT scheduler tick: 7 Number of interfaces: 3 Number of LSAs in DB: 308 Number of neighbors: 1 Number of adjacent neighbors: 1 bird> show ospf nei MyOSPF: Router ID Pri State DTime Router IP Interface 147.251.240.166 1 full/dr 00:34 147.251.61.1 eth0 bird> show ospf int MyOSPF: Interface "eth0": Area: 0.0.0.2 (2) Type: broadcast State: backup Priority: 1 Cost: 500 Hello timer: 10 Wait timer: 40 Dead timer: 40 Retransmit timer: 5 Designed router (ID): 147.251.240.166 Designed router (IP): 147.251.61.1 Backup designed router (ID): 147.251.5.1 Backup designed router (IP): 147.251.61.2 Interface "eth1": Area: 0.0.0.2 (2) Type: broadcast State: dr Priority: 1 Cost: 500 Hello timer: 10 Wait timer: 40 Dead timer: 40 Retransmit timer: 5 Designed router (ID): 147.251.5.1 Designed router (IP): 147.251.61.6 Backup designed router (ID): 0.0.0.0 Backup designed router (IP): 0.0.0.0 Interface "eth2": Area: 0.0.0.2 (2) Type: broadcast State: dr (stub) Priority: 1 Cost: 10 Hello timer: 10 Wait timer: 40 Dead timer: 40 Retransmit timer: 5 Designed router (ID): 147.251.5.1 Designed router (IP): 147.251.5.1 Backup designed router (ID): 0.0.0.0 Backup designed router (IP): 0.0.0.0 ---------------- I could not see any coincidence with any event :-( I have got HP Internet Advisor, so i can debug ospf well, but this internal network is production network and I can not play with it. ---------------- Another problem: I accidentally run two bird daemons. When I killed the second one, I could not connect to /var/bird.ctl. Even without looking to source, i guess, that bird doesn't check existence of other bird daemon and after on exit unlinks /var/bird.ctl. :-( I know, running two same daemons is stupid idea, but it was a typo, i wanted to type birdc :-( -- ------------------------------------------------------------------------- David Rohleder davro(a)ics.muni.cz Institute of Computer Science, Masaryk University Brno, Czech Republic -------------------------------------------------------------------------

1 0

problems with simple configuration
by David Rohleder 09 Apr '01

09 Apr '01

Hello, i have got a simple config, which behaves badly: Internet P-t-P netw. local network ------------| VNR |----------| R |------------- ---[/etc/bird.conf debug protocols all; protocol direct { # interface "-eth*", "*"; # Restrict network interfaces it works with } protocol kernel { # learn; # Learn all alien routes from the kernel persist; # Don't remove routes on bird shutdown scan time 20; # Scan kernel routing table every 20 seconds # import none; # Default is import all export all; # Default is export none # kernel table 5; # Kernel table to synchronize with (default: mai n) } protocol static { } protocol rip MyRIP { # You can also use an explicit name port 520; interface "eth2" { mode broadcast; }; authentication none; import none; export all; } protocol ospf MyOSPF { # rfc1583compat yes; area 0.0.0.2 { # stub no; # tick 1; interface "eth0" { cost 500; dead count 4; type broadcast; }; interface "eth1" { cost 500; dead count 4; type broadcast; }; interface "eth2" { stub yes; cost 10; }; interface "eth3" { stub yes; cost 10; }; }; } ------[/etc/bird.conf eth1 and eth3 are up, but there are not any machines. eth3 doesn't have IP address. After start, VNR (Bay Networks) doesn't see internal network. But they are in full state! And even more strange thing: VNR learns LSA after some action in birdc. For example, when i run debug MyOSPF all, or even just start birdc! May be some timing problems. It is PIII/600, 128MB RAM. some strange messages: MyOSPF: LS ack sent to 147.251.61.1 (delayed) shows every max. 20 sec. 147.251.61.0/30 is P-t-P link. I have got BIRD 1.0.4 with OSPF area patch. Please, release 1.0.5 as soon as possible. Or at least publish all patches. -- ------------------------------------------------------------------------- David Rohleder davro(a)ics.muni.cz Institute of Computer Science, Masaryk University Brno, Czech Republic -------------------------------------------------------------------------

2 1

Re: BIRD doesn't compile
by David Rohleder 06 Apr '01

06 Apr '01

"James A. Crippen" <james(a)UnLambda.COM> writes: > On Thu, 5 Apr 2001, Florian Lohoff wrote: > > > On Thu, Apr 05, 2001 at 03:48:34PM +0200, David Rohleder wrote: > > > > > > if [ -S /var/bird.ctl ] ; then > > > echo already running > > > fi > > > > > > does the same thing. > > > > > And off course many daemons are using start-stop-daemon > > > > Just by staring at it i saw a couple of ugly things - Must have been > > very early when i actually did it. > > > > The above would only imply that the pipe exists but not actually > > tell you if bird is running. > > > > "birdc" exits with an error code > 0 if bird is not reachable which > > i try to detect. > > > > The problem with the start-stop-daemon is that it needs a pid file > > which when i remember correctly is not written by bird. One > > solution would have been to make a patch to bird (Which i had > > done if i remember) but i dropped it because of lazyness. > > I have my own init system which uses a C program similar to debian's > start-stop-daemon. I've written it such that given a command line flag > the daemon spawning program will write a pid file if you know the daemon > you're running doesn't write one (such as with bird). Once I did this I > had no trouble starting and stopping bird using the pid file > mechanism. Perhaps you might patch debian's start-stop-daemon to do a > similar thing? It's a thought. start-stop-daemon already does this: START-STOP-DAEMON(8) Debian GNU/Linux START-STOP-DAEMON(8) --make-pidfile Used when starting a program that does not create its own pid file. This option will make start-stop- daemon create the file referenced with --pidfile and place the pid into it just before executing the process. Note, it will not be removed when stopping the program. NOTE: This feature may not work in all cases. Most notably when the program being exe cuted forks from it's main process. Because of this it is usually only useful when combined with the --background option. -- ------------------------------------------------------------------------- David Rohleder davro(a)ics.muni.cz Institute of Computer Science, Masaryk University Brno, Czech Republic -------------------------------------------------------------------------

3 2

Re: BIRD doesn't compile
by David Rohleder 05 Apr '01

05 Apr '01

Florian Lohoff <flo(a)rfc822.org> writes: > On Thu, Apr 05, 2001 at 02:29:52PM +0200, David Rohleder wrote: > > > > I was looking for debian package, but i was not able to find it. I was > > quite happy with that, because I wanted to become a debian > > developper. Bad luck this time. I will have to look for another > > package. > > > You are using interesting init.d script :-) I have never seen this before. > > You mean this ? > > if [ -d /proc -a -z "`cat /proc/ksyms | grep rtnetlink`" ]; then > > I wanted to give a resonable output if people dont compile > in "rtnetlink" - I mean if they even skip proc there is no hope. No, I mean this kind of starting and other operations with bird ---- echo quit | birdc >/dev/null 2>&1 if [ $? -eq 0 ]; then echo already running fi ---- if [ -S /var/bird.ctl ] ; then echo already running fi does the same thing. --- And off course many daemons are using start-stop-daemon -- ------------------------------------------------------------------------- David Rohleder davro(a)ics.muni.cz Institute of Computer Science, Masaryk University Brno, Czech Republic -------------------------------------------------------------------------

3 2

Re: BIRD doesn't compile
by David Rohleder 05 Apr '01

05 Apr '01

Florian Lohoff <flo(a)rfc822.org> writes: > On Tue, Mar 06, 2001 at 10:49:18AM +0100, David Rohleder wrote: > > Hello, > > > > i have got a problem compiling BIRD 1.0.4 on newly installed debian > > box with gcc 2.95.3 > > apt-get install bird ? Doesn't work with OSPF area other than backbone. (I posted a patch month ago). I was looking for debian package, but i was not able to find it. I was quite happy with that, because I wanted to become a debian developper. Bad luck this time. I will have to look for another package. You are using interesting init.d script :-) I have never seen this before. > Flo > -- > Florian Lohoff flo(a)rfc822.org +49-5201-669912 > Why is it called "common sense" when nobody seems to have any? -- ------------------------------------------------------------------------- David Rohleder davro(a)ics.muni.cz Institute of Computer Science, Masaryk University Brno, Czech Republic -------------------------------------------------------------------------

3 2

Re: BIRD doesn't compile
by David Rohleder 05 Apr '01

05 Apr '01

Ondrej Feela Filip <feela(a)ipex.cz> writes: > > I was looking for debian package, but i was not able to find it. I was > > quite happy with that, because I wanted to become a debian > > developper. Bad luck this time. I will have to look for another > > package. > > You can maintain a debian package of Bird. Superb, but bird is not orphaned package and Florian Lohoff is still it's maintainer. -- ------------------------------------------------------------------------- David Rohleder davro(a)ics.muni.cz Institute of Computer Science, Masaryk University Brno, Czech Republic -------------------------------------------------------------------------

1 0

IS-IS ?
by Florian Lohoff 05 Apr '01

05 Apr '01

Hi, some people at work think that replacing eigrp by IS-IS is "a good think" (tm). Is there anyone working on IS-IS support for Bird ? Flo -- Florian Lohoff flo(a)rfc822.org +49-5201-669912 Why is it called "common sense" when nobody seems to have any?

1 0

BIRD doesn't compile
by David Rohleder 05 Apr '01

05 Apr '01

Hello, i have got a problem compiling BIRD 1.0.4 on newly installed debian box with gcc 2.95.3 gcc 2.95.2 compiles without errors (but with many warnings). Here is what i have got (seems that #include <time.h> is missing) on 2.95.3 io.c: At top level: io.c:438: warning: missing initializer io.c:438: warning: (near initialization for `sk_class.lookup') io.c: In function `io_init': io.c:1021: warning: implicit declaration of function `time' make[2]: *** [io.o] Error 1 make[2]: Leaving directory `/home/davro/source/bird-1.0.4/obj/lib' make[1]: *** [subdir] Error 2 make[1]: Leaving directory `/home/davro/source/bird-1.0.4/obj' make: *** [all] Error 2 when i added #include <time.h> to several sources, it has passed without errors. But i don't know if this is "the right way" to do it. -- ------------------------------------------------------------------------- David Rohleder davro(a)ics.muni.cz Institute of Computer Science, Masaryk University Brno, Czech Republic -------------------------------------------------------------------------

3 2

Is this an active project?
by Ralph Jennings 02 Apr '01

02 Apr '01

Is this an active project? There seems to be no changes since September 2000? Are all the developers on vacation? Permanently? -Ralph

4 4

About func in md5.c
by Zheng Yuan {ZYUAN1} 07 Mar '01

07 Mar '01

Hi, All maybe i totally dont figure this right, in the func void byteReverse(unsigned char *buf, unsigned longs) { u32 t; do { t = (u32) ((unsigned) buf[3] << 8 | buf[2]) << 16 | ((unsigned) buf[1] << 8 | buf[0]); *(u32 *) buf = t; buf += 4; } while (--longs); } isnt that buf[3] should shift 24?

1 0