[c-nsp] Cisco CEF "Encap_fail"

[Rodney Dunn]

We record an Encap_fail for a stalled adjacency. That means
that we punted the first packet and arp'd for it.
While waiting on the response any subsequent packets are dropped
by rate limiting to the processor.

You can turn on: debug ip cef drop and see 'em.

Or check your arp table for incomplete entries is the most common
place to look.

On Fri, Oct 28, 2005 at 12:43:29PM -0500, Brad Swanson wrote:
> I have the following in the RP of a 6509, sup1a/MSFC2.
> 
> #sh cef drop
> 
> Slot  Encap_fail      Unresolved Unsupported    No_route      No_adj  
> ChkSum_Err
> RP     265238851           0            0              9275519           0   
>          0
> 
> The Encap_fail creeps up about 4-5 every 10 seconds.  Any ideas?  Something 
> to worry about?  The processor runs at 2-3%.  I also have Input queue 
> drops...and have a case open with Cisco on this, just need to get a sniffer 
> trace...
> 
> 20 seconds after clearing counters:
> Input queue: 1/75/56/0 (size/max/drops/flushes); Total output drops: 0
> 20 seconds or so later I have 72 drops.
> 
> Different animals or is there a relationship between the two?  I understand 
> CEF and queues, just want to know if there is any way the two issues are 
> related.
> 
> Thanks,
> Brad Swanson
> 
> 
> _______________________________________________
> 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/