Problem With E1

Graham, Darel Darel.Graham at Globalcrossing.com
Thu Feb 26 12:50:39 UTC 2009


Using an external CSU? Power cycle it. 
Using Frame Relay? Have provider check card in Frame switch, if the port reads 65535 the buffers are full. Have the provider reset the card (shared memory cards are bad). 

Without configuration info or other supporting info it will be difficult to tell what the underlying issues are with OSPF/Ckt. Thanks Howard for the push in the right direction.

DR Graham
-----Original Message-----
From: Howard C. Berkowitz [mailto:hcb at netcases.net] 
Sent: Thursday, February 26, 2009 7:37 AM
To: nanog at nanog.org
Subject: RE: Problem With E1



> -----Original Message-----
> From: Shivlu Jain [mailto:shivlu.jain at gmail.com]
> Sent: Thursday, February 26, 2009 4:05 AM
> To: nanog at nanog.org
> Subject: Problem With E1
> 
> Since morning I am facing a issue in which one of E1 is configured under
> OSPF. OSPF neighborship is up but not able to send and receive the data.
> The
> configuration is plain vanila. Why it is happening so; I donot know?
> 
> --
> Thanks & Regards
> shivlu jain
> http://shivlu.blogspot.com/
> 09312010137

If  this is an operational circuit, this is  a good example of why it can
extremely useful to document the working configuration of a resource, so you
can compare the malfunctioning configuration. The document may well be
stored as a file, and the comparison could be made with diff or a similar
utility.

Don't forget SNMP and NetFlow, both on the router, but also SNMP on the
access device, modem, multiplexer, etc. 

When that circuit first came up, I probably would have captured the
information from the router's equivalent of the Cisco commands:

* show interface
* show ip interface
* show ip ospf interface
* show ip ospf neigbors

Possibly show ip ospf database  and show ip ospf database neighbors; perhaps
save the routing table when storing those displays.

Even more displays could be useful, such as subinterfaces. 

Electrical tests, such as verifying the signal clocking and amplitude, are
usually last resorts -- although do verify that no one has moved the cabling
among router/CSU ports, and that everything has power.






More information about the NANOG mailing list