[c-nsp] Cisco 6506E 4Byte ASN IOS Support

Nick Cutting ncutting at edgetg.co.uk
Mon Apr 11 06:39:20 EDT 2016


Doesn't the OP have a 65k with a sup720 - not a Sup2T/65/68k like everyone is replying with advice on?

-----Original Message-----
From: cisco-nsp [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Mark Tinka
Sent: 08 April 2016 14:59
To: Gert Doering; Nick Hilliard
Cc: cisco-nsp at puck.nether.net
Subject: Re: [c-nsp] Cisco 6506E 4Byte ASN IOS Support



On 8/Apr/16 12:23, Gert Doering wrote:

> OTOH we don't do "truly full" routes on the XL boxes either (stripped 
> down by a lot of /23 and /24 deaggregates from other regions), so they 
> are still mostly usable - wouldn't run a "full full!" table on Sup720s 
> anymore today.  BGP scheduler is too bad for the slowness of the CPU 
> (100% CPU is not something I worry about - that's what the CPU is 
> there for :-) - but 100% CPU due to one BGP neighbour sending in tons 
> of updates should never *ever* cause flapping on other BGP neighbours 
> due to missed keepalives)...

When we used to run XR 12000's back in the day under IOS 12.0, I was amazed by this.

The CPU ran at 100% and logged "CPU hog" forever, but neither IS-IS, BGP nor LDP sessions (90% of which had full BGP feeds) ever dropped.

Still amazes me today, especially when pings and SSH access to the routers was sluggish as mad...

Mark.



More information about the cisco-nsp mailing list