Thanks for the input all. It appears unanimous: My TCAM is stuffed. I'm
a little baffled by the EIGRP aspect (which I don't think anyone
addressed), but it makes sense that it would all be using the same
resources.
Is there not a simple command to show the used capacity of the TCAM?
Thanks!
--Rick
On Thu, 13 Aug 2009, Jon Lewis wrote:
On Thu, 13 Aug 2009, Jared Mauch wrote:
Your tcam is full. It may not log that, you likely need 12.2sxe or sxf to
see the logs on sup2. Cut down the number of routes you listen to from your
upstreams, having them send you just their customer routes and use default
for the rest.
In sup720 and sup1a when the tcam is full it would then process switch. Not
sure on sup2 but I presume it's the same.
When we were discussing this on-list a year or more ago, I think someone said
that what the sup2 did when tcam filled was IOS version dependent. Newer IOS
would process switch. Older IOS would blackhole. I never verified this.
I really expected to see messages like this about a year ago. A full view is
around 290k routes...way more than the sup2 tcam can handle. This guy has to
have been having issues for months.
----------------------------------------------------------------------
Jon Lewis | I route
Senior Network Engineer | therefore you are
Atlantic Net |
_________ http://www.lewis.org/~jlewis/pgp for PGP public key_________
_______________________________________________
cisco-nsp mailing list [email protected]
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
