OSPF over direct fiber link and backup link

[enriquebs]

HI,

In this scenario there are no problems, is just that we need a justification about a possible issue in the backup network of our Service Provider.

RouterPE-A ----- direct fiber link ----- RouterP-X

RouterPE-A ----- backup link ----- RouterP-Y

The Service provider give us connectivity over a direct fiber link and also from a MetroEthernet backup link between a PE-Router and two P-Routers (MPLS environment). We lost OSPF and BGP adjacencies at the same time in the both links while interfaces are always up. We think the problem could be in a common device between the routers, maybe a JDS box. The thing is, how could we discart that the problem is in our PE router in order to give to our service provider the tests? We wanna discart a protocol issue in our PE, have anyone any idea about a debug command or something like that to get that information?

Thanks a lot in advance.

Regards,

[jbrunner007]

you should look into ospf bfd enhancements...

check out

http://www.cisco.com/en/US/products/ps6599/products_white_paper0900aecd80244005.shtml

bfd should allow you accurately determine true reachability between ospf neighbors.. another thing you can do is "debug ip ospf adj" this will tell why things are happening...

BTW: dont you mean "CE"? PE is the providers edge box, whereas CE is your customer edge box?

Or are you doing inter-carrier MPLS, and you are a provider?

Joe

[refayet.ullah]

Hi,

We are experiencing same problem for our Core routers in WAN. Which kind of interface you are using (electrical FE) to connect to the SDH MUX.

In our case, SDH MUX can not send any notification to the WAN interface when the SDH cross connect (VC4) is broaken. Our vendor (SIEMENS) is working on any patch.

Thats why, our FE WAN interfaces is up but can not get any OSPF hello from neibour. You can try OSPF fast hello for faster convergence.

Regards,

Refayet