Intermittent Latency Issues

[SrimantaM]

Hi Guys,

I am experiencing high latency intermittently in our internal LAN: Please help to resolve:::::

I have configured two 5940 switches in stack. 4*40Gbps ports are used for irf. Stacking is working fine. Three  "HPE OEM ProLiant BL460c Gen9"blade server are integrated with these switch. Six BAGG interfaces are created to integrate the server. Each BAGG interface contains two 40Gbps port(One from switch1 & another from Switch2). All gateways are defined in this switch. When I pinging Server IP from switch getting high latency intermittently for some server IP. If I shut switch-1' port(used in BAGG) getting latency for all server IP. But when shutting switch-2' port(used in BAGG) no latency for all IP.

Means when packet is coming via second switch, getting latency.

I checked by interchanging the cable but still latency is there. Same issue is observed for all six BAGG interface.

IP 10.60.205.1/24 is configured in this switch on interface vlan600. ping response for 10.60.205.16 ::

56 bytes from 10.60.205.16: icmp_seq=16 ttl=64 time=1.516 ms
56 bytes from 10.60.205.16: icmp_seq=17 ttl=64 time=1.464 ms
56 bytes from 10.60.205.16: icmp_seq=18 ttl=64 time=72.630 ms
56 bytes from 10.60.205.16: icmp_seq=19 ttl=64 time=19.773 ms
56 bytes from 10.60.205.16: icmp_seq=20 ttl=64 time=62.238 ms
56 bytes from 10.60.205.16: icmp_seq=21 ttl=64 time=2.343 ms
56 bytes from 10.60.205.16: icmp_seq=22 ttl=64 time=1.600 ms
56 bytes from 10.60.205.16: icmp_seq=23 ttl=64 time=2.292 ms
56 bytes from 10.60.205.16: icmp_seq=24 ttl=64 time=1.522 ms
56 bytes from 10.60.205.16: icmp_seq=25 ttl=64 time=20.496 ms
56 bytes from 10.60.205.16: icmp_seq=26 ttl=64 time=44.124 ms
56 bytes from 10.60.205.16: icmp_seq=27 ttl=64 time=2.002 ms

No latency from 10.60.205.13

56 bytes from 10.60.205.13: icmp_seq=0 ttl=64 time=1.468 ms
56 bytes from 10.60.205.15: icmp_seq=1 ttl=64 time=1.359 ms
56 bytes from 10.60.205.13: icmp_seq=2 ttl=64 time=1.493 ms
56 bytes from 10.60.205.13: icmp_seq=3 ttl=64 time=1.185 ms
56 bytes from 10.60.205.13: icmp_seq=4 ttl=64 time=1.156 ms
56 bytes from 10.60.205.13: icmp_seq=5 ttl=64 time=1.166 ms
56 bytes from 10.60.205.13: icmp_seq=6 ttl=64 time=1.266 ms
56 bytes from 10.60.205.13: icmp_seq=7 ttl=64 time=1.309 ms
56 bytes from 10.60.205.13: icmp_seq=8 ttl=64 time=1.226 ms
56 bytes from 10.60.205.13: icmp_seq=9 ttl=64 time=1.096 ms

<TOR01>display version

HPE Comware Software, Version 7.1.070, Release 2612P02

Copyright (c) 2010-2018 Hewlett Packard Enterprise Development LP

HPE FF 5940 32QSFP+ Switch uptime is 0 weeks, 0 days, 5 hours, 49 minutes

BAG Config::

========

interface Bridge-Aggregation24

description To_Enc_2_VC1_Q1_Q2_External

port link-type trunk

undo port trunk permit vlan 1

port trunk permit vlan 600 to 699

link-aggregation mode dynamic

TEST CASE::

In BAGG 24 port 1/0/19 & 2/0/19 is there. If shut 1/0/9 then latency for all IP. If shut 2/0/9 no latency for all IP. If plug out the cable from 2/0/9 and connect to 1/0/9 no latency is there.

Switch reboot and upgradation is also done but no luck.

[jmpk]

Hello,

As per my understanding , when you shutdown IRF primary interface you are facing the latencey for All IP  but when you shutdown secondary interface you not seeing latency . This is an expected behaviour for ICMP traffic in IRF , when traffic pass thru IRF primary will not have any latencey . 

Thanks

JMPK