[nsp] 6500 Native IOS ACLs

[Rubens Kuhl Jr.]

Unresolved adjacencies that happens to have lots of traffic targeted at them
also pump up MSFC CPU load...  PFC cannot resolve adjacencies by itself,
relying on punting packets thru MSFC and get a complete adjacency after some
time.




Rubens Kuhl Jr.


----- Original Message -----
| router#show ip cef sum
| IP Distributed CEF with switching (Table Version 7169682), flags=0x0
|   120239 routes, 0 reresolve, 120 unresolved (94 old, 26 new), peak 116897
|   120239 leaves, 8114 nodes, 22887456 bytes, 7143003 inserts, 7022764
invalidations
|   321 load sharing elements, 107856 bytes, 120028 references
|   universal per-destination load sharing algorithm, id 9078CCCE
|   3 CEF resets, 35036 revisions of existing leaves
|   Resolution Timer: Exponential (currently 4s, peak 16s)
|   26691 in-place/0 aborted modifications
|   refcounts:  1853310 leaf, 1677246 node
|
| Adjacency Table has 55 adjacencies
|
|
|
| On Wed, Nov 13, 2002 at 04:19:03PM -0200, Rubens Kuhl Jr. wrote:
| >
| > Is RPF checking enabled on any interface ? This halves maximum FIB size
to
| > 112,000 routes... replacing RPF with inbound ACLs removes this
limitation.
| > Even if it's not your current problem, it will haunt you very soon
unless
| > you filter a lot of routes and default them.
| >
| >
| > Rubens Kuhl Jr.
| >
|
| --
| =========================================================================
| Clinton Work                                        clinton@scripty.com
| Calgary, Alberta
| _______________________________________________
| cisco-nsp mailing list  real_name)s@puck.nether.net
| http://puck.nether.net/mailman/listinfo/cisco-nsp
| archive at http://puck.nether.net/pipermail/cisco-nsp/