[j-nsp] ispf support on Juniper routers

Mark Tinka mtinka at globaltransit.net
Tue Mar 10 00:36:07 EDT 2009


On Tuesday 10 March 2009 11:53:57 am Paul Goyette wrote:

> The other interesting thing is that the router needs to
> keep additional state information in the internal OSPF
> routing tables to enable ISPF.  In my opinion, with
> today's routers, memory seems to be more of a limitation
> than CPU horsepower.

But I imagine we'd be talking about control plane memory 
here, not switch fabric memory (FIB), where the former is 
more expandable (in relation, since they really aren't the 
same thing).

Besides, if networks follow the good practice of using IGP's 
to hold infrastructure and Loopback addresses only, and 
throw the rest into iBGP, the scalability of the IGP is far 
greater.

But I do agree, while iSPF and PRC help make CPU utilization 
more efficient, they actually mostly shine during periods 
where bad things are happening to the network, e.g. multiple 
link or node failures, control plane attacks, e.t.c. So 
under normal circumstances, today's routers have quick 
enough CPU's for SPF calculations not to be such a huge 
problem (good fundamental design being the case. of course).

Cheers,

Mark.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 835 bytes
Desc: This is a digitally signed message part.
URL: <https://puck.nether.net/pipermail/juniper-nsp/attachments/20090310/2c10d520/attachment.bin>


More information about the juniper-nsp mailing list