bug-zebra
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: zebra/ospfv3 help


From: Yasuhiro Ohara
Subject: Re: zebra/ospfv3 help
Date: Wed, 06 Feb 2002 02:26:30 +0900 (JST)

Mate.Csorba> Dear Colleagues!
Mate.Csorba> 
Mate.Csorba> We'd like to have some help from You regarding the ospf6d
Mate.Csorba> daemon included in the zebra0.92a package. We are runing
Mate.Csorba> OSPFv3 conformance tests on a local ethernet network and
Mate.Csorba> we're experiencing problems with the sending of Database
Mate.Csorba> Description packets. As You can see in the dump file
Mate.Csorba> included, the ospf6d daemon sends the DD packets
Mate.Csorba> required, but this packets won't be passed to the network
Mate.Csorba> somehow.
Mate.Csorba> They can not be detected even with some network
Mate.Csorba> monitoring software, although the sending and of Hello
Mate.Csorba> packets seems to be perfect (we receive them, and the
Mate.Csorba> monitor can see them too).

Hi, thank you for your testing.

What is the neighboring router ? The neighbor router sends Hellos
specifying "fe80::b" as the source address, which seems unusual.

2002/02/05 16:20:30 OSPF6: Receive Hello on eth0
2002/02/05 16:20:30 OSPF6:     fe80::b -> ff02::5
2002/02/05 16:20:30 OSPF6:     OSPFv3 Type:1 Len:40 RouterID:10.1.11.2
    :
    :

Remember, the Database Description packet is sent to the neighbors
address directly. So if the monitor's machine is not the same as the
neighbor OSPF software, monitor usually can't see Database packet.

If the neighbor OSPF router doesn't see the Database packet, check
whether the link-local address of that machine is "fe80::b".

I know there are (at least) two problems in ospf6d:
  - ospf6d crashes if it receives a LSA whose scope bits in LS-Type
    indicate unknown scope.
  - ospf6d fails to remove obsolete routes
I will fix this in a week or two.

If you find another problem, please please let me know.

regards.
yasu




reply via email to

[Prev in Thread] Current Thread [Next in Thread]