OSPF loading State

Ryan Whelan rcwhelan at gmail.com
Fri Jul 22 17:00:09 CEST 2011


On Fri, Jul 22, 2011 at 3:37 AM, Tapio Haapala
<tapio.haapala at f-solutions.fi> wrote:
> We have similar issues but looks that it get fixed when you take newest svn
> version. BUT you get new problem that some times some routes does not sync
> and you need restart bird =)

Alright, well as long as its a known bug i can work around it for now.
Still rather use bird than quagga if i can get away with it

>
> 21.7.2011 23:39, Ryan Whelan kirjoitti:
>>
>> On Thu, Jul 21, 2011 at 12:05 PM, Ryan Whelan<rcwhelan at gmail.com>  wrote:
>>>
>>> I have 2 linux machines connected to one another over the internet via
>>> a IPSec protected GRE tunnel.  Each of them are running OSPF from bird
>>> 1.3.2.
>>>
>>> The issue is that one of them will not fully peer with the other,
>>> unless the other is started after the one with the issue.  Is just
>>> stays in state 'loading' while the other machine will be in state
>>> 'full'.  In other words, node 2 will only go to state 'full' if node 1
>>> is started after node 2. If node 2 is started after node 1, it will
>>> sit in state 'loading' forever. Node 1 has no issues.
>>>
>>> Here is a log from the node that is failing to sync up (node 2):
>>> http://pastebin.com/ZjVcc1wB
>>>
>>>
>>> bird>  show ospf neighbors
>>> axa0:
>>> Router ID       Pri          State      DTime   Interface  Router IP
>>> 111.222.333.444   1      loading/ptp    00:13   tun10      172.31.255.255
>>> bird>
>>>
>>>
>>>
>>> the config on both machines is identical
>>>
>>>
>>>
>>> protocol kernel {
>>>       learn;;
>>>        persist;
>>>        scan time 20;
>>>        export all;
>>> }
>>>
>>> protocol device {
>>>        scan time 10;
>>> }
>>>
>>> protocol static {
>>> }
>>>
>>> protocol ospf axa0 {
>>>        tick 1;
>>>        rfc1583compat yes;
>>>        area 0.0.0.0 {
>>>                stub no;
>>>                interface "tun*" {
>>>                        hello 5;
>>>                        retransmit 5;
>>>                        cost 10;
>>>                        transmit delay 5;
>>>                        dead 15;
>>>                        wait 50;
>>>                        type ptp;
>>>                };
>>>       };
>>> }
>>>
>> If it means anything, the machine that was stuck in 'loading' changed
>> to 'full' after I left it alone, but it took about 30 minutes.  Is
>> that expected behaviour?
>
> --
> F-Solutions Oy
>
> Tapio Haapala
>
> PL 7, 90571 Oulu
> GSM   040-0998371
> Skype burner-
> IRC   Burner at ircnet
>
>
>



More information about the Bird-users mailing list