[c-nsp] Cisco 6500 packet loss

naresh reddy nareshbtech at yahoo.com
Wed Feb 26 10:24:34 EST 2014


Hi
 we don't manage remote end so cant verify the return path


this router dosen't hold full bgp table


total routes in this box is ~500 prefixes


the route being used for this destination is default route form isp



M(134368 ): E | 1 FFF  0 0 0 0   0.0.0.0
V(134368 ): 8 | 1 0    0 0 0 0   0.0.0.0            (A:16402  ,P:0,D:0,m:0 ,B:0 )


sh mls cef adjacency entry 16402 detail
Index: 16402   smac: 0015.c7a0.8c00, dmac: e8b7.4829.eac0
               mtu: 1518, vlan: 1022, dindex: 0x0, l3rw_vld: 1
               format: MAC_TCP, flags: 0x2000008408
               delta_seq: 0, delta_ack: 0
               packets: 174350, bytes: 33339216


 
Nareshchandra Paturi





On Wednesday, February 26, 2014 2:44 PM, Saku Ytti <saku at ytti.fi> wrote:
 
On (2014-02-26 06:30 -0800), naresh reddy wrote:

> svis are on the 6500 destination is 11 hops away and route is learned from BGP table

Lot of domain to cover here, 11 hops :)

Can you verify return path for working and non-working? Are they using same
path or different path?
Can you verify which direction packet loss is (if you have PC in both
directions, it's easy with iperf or so)

If you manage the path end-to-end, and if it is 7600/6500 you can try to check
path in HW via:

sh mls cef lookup DST detail
sh mls cef adjacency entry XYZ detail (XYZ from above output A:12345)

If you have DFC (in device you showed you don't have), you need to check for
ingress card DFC by prepending command with 'module X'

-- 
  ++ytti

_______________________________________________
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/


More information about the cisco-nsp mailing list