[c-nsp] BGP default vs BGP full

[Brian Wallingford]

Still, no need for bgp in a single-upstream environment.  You can use a dual default, so long as the upstream-facing interface physically (not just logically) goes down, which will automatically remove the route.  If there is an intermediate device between the fiber drop and your router, chances are the ethernet will appear up even if connectivity is not, which will be problematic as the route will not disappear.  In that case, look into IP sla.

Sent from my iPad

> On May 20, 2016, at 10:02 PM, Satish Patel <satish.txt at gmail.com> wrote:
> 
> Currently we have single link but in sure if i get second fiber drop
> from same ISP then can i use it for loadbalacing or failover? In that
> case what kind of BGP attributes i need to specify?
> 
>> On Fri, May 20, 2016 at 6:06 PM, Brian Wallingford <brian at meganet.net> wrote:
>> If you only have a single egress, there's no real advantage to taking a full table unless you are interested in using it as a learning tool.  By simply using a default, you'll save loads of memory, as well as CPU should you drop the connection during, e.g., maintenance.
>> 
>> Cheers,
>> Brian
>> 
>>> On May 20, 2016, at 5:14 PM, Satish Patel <satish.txt at gmail.com> wrote:
>>> 
>>> Just wonder what would be the advantage and disadvantage of running
>>> BGP full vs default route.
>>> 
>>> We have single ISP connection and ISP decided to just run default
>>> route over BGP instead full.
>>> _______________________________________________
>>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>>> archive at http://puck.nether.net/pipermail/cisco-nsp/
>>