Re: ospfd errors

2007-11-14 Thread Esben Norby
On Wednesday 14 November 2007 06:11:43 Linden Varley wrote:

 I was wondering if anyone could offer any solution to this OSPFD error
 when it starts up:

 ospfd[11601]: send_packet: error sending packet on interface em0: No
 route to host

 It says there is no route to host for every interface defined in ospfd.conf

 This is using the default config on an OpenBSD 4.0 amd64 install.

Please post your /etc/ospfd.conf and the output of ifconfig here.

OpenBSD 4.0 is old, could you try with something newer?

 (Note, ip forwarding and ip multicast forwarding are enabled)

No need to enable ip multicast forwarding.

/Esben



Re: ospfd: some machines stuck in 2-WAY/OTHER

2007-08-20 Thread Esben Norby
On Monday 20 August 2007 12:37:03 Toni Mueller wrote:
 Hi,

 in my network, I have five OSPF routers and hosts which learn their routes
 via OSPF, all in one area. One of the routers is a Cisco, all other
 affected routers and hosts are OpenBSD/i386 4.1-stable as of May 24th. The
 problem is that some machines establish an adjacency with some, but not all
 machines in that area.

I think you network is just fine.

From the output I gather that the three routers all agree that 10.0.0.6 is the
DR and that the 10.0.0.4 is the BACKUP.




This is 10.0.0.4 - it has FULL with all routers and the DR.

 On 192.168.50.4:

 $ ospfctl show nei
 ID  Pri StateDeadTime Address Iface Uptime
 10.0.0.25   FULL/OTHER   00:00:37 192.168.50.2   fxp0  00:19:04
 10.0.0.550  FULL/OTHER   00:00:37 192.168.50.5   fxp0  01w5d17h
 10.0.0.650  FULL/DR  00:00:33 192.168.50.6   fxp0  01w5d17h
 10.0.0.35   FULL/OTHER   00:00:34 192.168.50.3   fxp0  02w2d12h
 10.0.0.110  FULL/OTHER   00:00:34 192.168.50.1   fxp0  01w6d21h


This is just a normal router thus it is FULL with the DR and the BACKUP,
ignoring the rest...


 On 192.168.50.3:

 $ ospfctl show nei
 ID  Pri StateDeadTime Address Iface Uptime
 10.0.0.25   2-WAY/OTHER  00:00:31 192.168.50.2   fxp1  -
 10.0.0.550  2-WAY/OTHER  00:00:31 192.168.50.5   fxp1  -
 10.0.0.650  FULL/DR  00:00:38 192.168.50.6   fxp1  01w5d17h
 10.0.0.110  2-WAY/OTHER  00:00:39 192.168.50.1   fxp1  -
 10.0.0.450  FULL/BCKUP   00:00:39 192.168.50.4   fxp1  02w2d12h



Another normal router that is FULL with the DR and the BACKUP - ignoring the
rest...


 On 192.168.50.2:

 $ ospfctl show nei
 ID  Pri StateDeadTime Address Iface Uptime
 10.0.0.550  2-WAY/OTHER  00:00:32 192.168.50.5   dc2   -
 10.0.0.110  2-WAY/OTHER  00:00:30 192.168.50.1   dc2   -
 10.0.0.450  FULL/BCKUP   00:00:39 192.168.50.4   dc2   00:19:04
 10.0.0.35   2-WAY/OTHER  00:00:39 192.168.50.3   dc2   -
 10.0.0.650  FULL/DR  00:00:38 192.168.50.6   dc2   00:19:04


 The router 192.168.50.1 is the Cisco machine.


HTH
Esben



Re: OSPF and IPv6

2007-02-28 Thread Esben Norby
On Wednesday 28 February 2007 14:58:49 Jon Morby wrote:
 Unless I'm missing something OpenOSPFD doesn't currently seem to
 support IPv6 ?

IPv6 is not supported currently, and I think it will be a while before that 
happens.

/Esben



Re: ospfd participating in a stub area

2007-02-07 Thread Esben Norby
On Wednesday 07 February 2007 00:53, Nigel Roberts wrote:
 Is it possible to configure an area in ospfd.conf to be a stub area?

 I have an area where all particpating routers (ciscos) are configured
 to treat it as a stub ie.

 router ospf 1234
 ...
   area 1 stub
 ...

What excactly is the purpose of this? Is it some cisco trick to save memory or 
does it have a real purpose?

Normaly when routers form adjacency the network is not considered a stub 
network any more, hence it can be used to forward traffic.

/Esben



Re: ospfd.conf example inheritance problem

2006-04-13 Thread Esben Norby
On Thursday 13 April 2006 00:38, andrew fresh wrote:
 I was trying to set default auth-type and auth-md and ran into some
 trouble.  Doing some debugging, I tried just uncommenting part of the
 example ospfd.conf and have found it doesn't work.  Here is what I did:


Oh yes indeed - I think you found a bug. The man page claims that your 
configuration is valid, but it doesn't seem to work. I will look into this.

/Esben



Re: OSPFD - No buffer available

2006-03-31 Thread Esben Norby
On Friday 31 March 2006 14:19, Anderson Nadal wrote:
 Hello.

 I'm using OpenBSD 3.8 and OSPFD  and carp.


If your using the ospfd that is in 3.8 I would reccommend updating 
ospfd/ospfctl to current. A lot of improvements has been made the last 10 
months. Amoung them is the handling of buffers and carp.

/Esben



Re: OSPFD - No buffer available

2006-03-31 Thread Esben Norby
On Friday 31 March 2006 18:45, Anderson Nadal wrote:
 Hi.

 I update ospfd from CVS (OPENBSD_3.9), and i have the same error.


Could you mail your ospfd.conf, and a trace from ospfd -d and ifconfig -a.

I normally see the no buffer space availible, when trying to transmit 
packets on an interface where the link has gone down. That however is 
normally prevented by the fact that ospfd senses link changes.

/Esben



Re: OpenBSD { future=PIM (DM-SM) } support or { only=XORP } ?

2006-02-07 Thread Esben Norby
On Tuesday 07 February 2006 01:56, Jason Houx wrote:

 I only read the protocol and never tried to set it up on a Crisco but now
 that the network is up I see no reason not to as I am not that interested
 in trying out XORP and can patiently hold my breath till I start to catch
 wind of some commits on the CVS posts.


If you manage to convince XORP to do PIM please let me know, that might come 
in handy when trying to produce some code for a OpenPIMD project.



Re: OpenBSD { future=PIM (DM-SM) } support or { only=XORP } ?

2006-02-05 Thread Esben Norby
On Friday 03 February 2006 18:12, Jason Houx wrote:

 I know you guys are busy working on what will become OpenBSD 3.9 so in no
 way do I mean to distract, but is anyone working on a OpenPIMd concept on
 the side (like you guys really need more side projects) or is there any
 hope for a OpenPIMd in the future (similar to a OpenOSPFd and OpenBGPd).

Yes, tiny baby steps has been taking in order to startup a OpenPIMD project, 
but don't hold your breath...

Step one would probably be a PIM-DM, later on it can be expanded to support 
PIM-SM.

If any one are interested in this sort of thing please chip in.

 I'm going to start playing with DVMRP now that I see I can use mrouted
 with our current PIM routers and hopefull get a tunnel up to transport the
 multicast traffic to a OpenBSD gateway.  This doesn't create a PIM router
 but it does give me more features to learn/play with on OpenBSD which is
 always a goal of mine.


I use mrouted myself on OpenBSD routers, and I really really want to replace 
them.

If you want to play, mrouted is good enough - I mainly use it for multicast 
video streaming, and IPerf multicast test streams.

/Esben



Re: Cryptographic authentication ospfd not working?

2005-10-28 Thread Esben Norby
On Friday 28 October 2005 09:15, Egbert Krook wrote:
 Hi,

 Does anyone know if MD5 authentication in ospfd is known to be broken? In
 our test environment things work fine without authentication or simple
 authentication turned on, but as soon as we turn on MD5 authentication
 things break.

 We're using the snapshot of October 13th and a Cisco 3640 router (IOS
 12.1(5)). If this is an unknown problem I will submit a bug report.


I cannot reproduce this with ospfd from current and a cisco 72xx router, using 
your configuration...

Could you provide a tcpdump trace?

/Esben



Re: ospf/gre or bgp over ipsec instead of cisco?

2005-05-05 Thread Esben Norby
On Tuesday 03 May 2005 23:47, Stephen Marley wrote:
 Can anyone forsee any problems doing this with the current state of ospf
 within openbsd? Or gif instead of gre perhaps?

One of the things that usally comes back and bites you is the lack of 
multicast support (or buggy multicast support). All (well almost all) OSPF 
signalling is sent as multicast traffic. A gre tunnel should handle that 
fine, haven't tried it but it seems it's just a point-to-point device.

The currnet state of ospfd should be enough to test your setup, and feedback 
from such a test would be greatly appreciated.

/Esben