07-14-2015 03:41 AM - edited 03-08-2019 12:57 AM
Hello, I have two N7706 with 48-port modules in slots 2 + 5, and have build a L3 port-channel using ports e2/1 and e5/1 each. Ports are connected with new SFP-H10GB-CU3M cables.
ping'ing across this port-channel shows 0.3 - 50% packet loss (depending on packet size). Any obvious errors in my config, or any explanation for this packet loss ? (loss pattern is same if I shut either e2/1 or e5/1 on a side)
Best Rgds,
Andreas
PS: Systems just arrived, have no smartnet ID yet - and our distributor hasn't managed to get TAC involved yet..
N7706-A# ping 10.64.255.18 vrf default packet-size 1500 count 50
PING 10.64.255.18 (10.64.255.18): 1500 data bytes
1508 bytes from 10.64.255.18: icmp_seq=0 ttl=254 time=2.043 ms
1508 bytes from 10.64.255.18: icmp_seq=1 ttl=254 time=2.057 ms
1508 bytes from 10.64.255.18: icmp_seq=2 ttl=254 time=1.636 ms
1508 bytes from 10.64.255.18: icmp_seq=3 ttl=254 time=1.527 ms
1508 bytes from 10.64.255.18: icmp_seq=4 ttl=254 time=1.595 ms
1508 bytes from 10.64.255.18: icmp_seq=5 ttl=254 time=1.561 ms
1508 bytes from 10.64.255.18: icmp_seq=6 ttl=254 time=1.675 ms
1508 bytes from 10.64.255.18: icmp_seq=7 ttl=254 time=1.617 ms
1508 bytes from 10.64.255.18: icmp_seq=8 ttl=254 time=1.67 ms
1508 bytes from 10.64.255.18: icmp_seq=9 ttl=254 time=1.524 ms
Request 10 timed out
1508 bytes from 10.64.255.18: icmp_seq=11 ttl=254 time=1.729 ms
1508 bytes from 10.64.255.18: icmp_seq=12 ttl=254 time=1.581 ms
1508 bytes from 10.64.255.18: icmp_seq=13 ttl=254 time=1.604 ms
1508 bytes from 10.64.255.18: icmp_seq=14 ttl=254 time=1.567 ms
1508 bytes from 10.64.255.18: icmp_seq=15 ttl=254 time=1.575 ms
1508 bytes from 10.64.255.18: icmp_seq=16 ttl=254 time=2.126 ms
1508 bytes from 10.64.255.18: icmp_seq=17 ttl=254 time=1.679 ms
1508 bytes from 10.64.255.18: icmp_seq=18 ttl=254 time=1.58 ms
1508 bytes from 10.64.255.18: icmp_seq=19 ttl=254 time=1.527 ms
1508 bytes from 10.64.255.18: icmp_seq=20 ttl=254 time=1.523 ms
Request 21 timed out
1508 bytes from 10.64.255.18: icmp_seq=22 ttl=254 time=1.57 ms
1508 bytes from 10.64.255.18: icmp_seq=23 ttl=254 time=1.551 ms
1508 bytes from 10.64.255.18: icmp_seq=24 ttl=254 time=1.579 ms
1508 bytes from 10.64.255.18: icmp_seq=25 ttl=254 time=1.548 ms
1508 bytes from 10.64.255.18: icmp_seq=26 ttl=254 time=1.548 ms
1508 bytes from 10.64.255.18: icmp_seq=27 ttl=254 time=1.553 ms
1508 bytes from 10.64.255.18: icmp_seq=28 ttl=254 time=1.535 ms
1508 bytes from 10.64.255.18: icmp_seq=29 ttl=254 time=1.599 ms
1508 bytes from 10.64.255.18: icmp_seq=30 ttl=254 time=1.552 ms
1508 bytes from 10.64.255.18: icmp_seq=31 ttl=254 time=1.572 ms
Request 32 timed out
1508 bytes from 10.64.255.18: icmp_seq=33 ttl=254 time=1.588 ms
1508 bytes from 10.64.255.18: icmp_seq=34 ttl=254 time=1.701 ms
1508 bytes from 10.64.255.18: icmp_seq=35 ttl=254 time=1.681 ms
1508 bytes from 10.64.255.18: icmp_seq=36 ttl=254 time=1.624 ms
1508 bytes from 10.64.255.18: icmp_seq=37 ttl=254 time=1.623 ms
1508 bytes from 10.64.255.18: icmp_seq=38 ttl=254 time=1.667 ms
1508 bytes from 10.64.255.18: icmp_seq=39 ttl=254 time=1.691 ms
1508 bytes from 10.64.255.18: icmp_seq=40 ttl=254 time=1.618 ms
1508 bytes from 10.64.255.18: icmp_seq=41 ttl=254 time=1.616 ms
1508 bytes from 10.64.255.18: icmp_seq=42 ttl=254 time=1.629 ms
Request 43 timed out
1508 bytes from 10.64.255.18: icmp_seq=44 ttl=254 time=1.723 ms
1508 bytes from 10.64.255.18: icmp_seq=45 ttl=254 time=1.602 ms
1508 bytes from 10.64.255.18: icmp_seq=46 ttl=254 time=1.654 ms
1508 bytes from 10.64.255.18: icmp_seq=47 ttl=254 time=1.63 ms
1508 bytes from 10.64.255.18: icmp_seq=48 ttl=254 time=1.612 ms
1508 bytes from 10.64.255.18: icmp_seq=49 ttl=254 time=1.638 ms
--- 10.64.255.18 ping statistics ---
50 packets transmitted, 46 packets received, 8.00% packet loss
round-trip min/avg/max = 1.523/1.636/2.126 ms
07-14-2015 08:30 AM
Hi,
Do you have the same issue if you don't use Portchannel at all? Just use one 10gig layer-3 connection and test again.
HTH
07-14-2015 08:39 AM
Hello,
thanks for your reply - yes, this happens even with only one int in the portchannel, or with direct links.
I've just found the cause, we had "copp profile strict" in place which polices icmp down to a very low rate (130 kbps), much too less for the ping tests..
#show policy-map interface control-plane
class-map copp-system-p-class-monitoring (match-any)
match access-group name copp-system-p-acl-icmp
match access-group name copp-system-p-acl-icmp6
match access-group name copp-system-p-acl-mpls-oam
match access-group name copp-system-p-acl-traceroute
match access-group name copp-system-p-acl-http-response
match access-group name copp-system-p-acl-smtp-response
match access-group name copp-system-p-acl-http6-response
match access-group name copp-system-p-acl-smtp6-response
set cos 1
police cir 130 kbps bc 1500 ms
conform action: transmit
violate action: drop
module 1:
conformed 8293884 bytes,
5-min offered rate 12928 bytes/sec
peak rate 14206 bytes/sec at Tue Jul 14 03:13:25 2015
violated 499852 bytes,
5-min violate rate 782 bytes/sec
peak rate 867 bytes/sec at Tue Jul 14 03:14:24 2015
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide