Interal Network Packet loss

[riazay001]

Dear All,

Its been a couple of days that I am having packet losses in my internal Network. Whenever I ping from Switch-12 to Switch-3 I can see about 10 packet loss out of 100 but when I ping 8.8.8.8 from the same switch it does not give any packet loss. This packet loss comes for about 2 to 4 seconds per minute.

The port between the switches is Trunk and all switches are publically accessible. There is no counter error, port errors etc in the Trunk ports.

Please advice.

[johnd2310]

HI,

Could be the switch you are pinging is busy.When a switch is very busy, processing pings is given low priority. What is the utilisation on that switch? 

Thanks

John

[riazay001]

Mr Johnd2310,

The Switch's cpu utilization is almost about 50 to 60 %.

[johnd2310]

Hi,

What is the output of the "show processes cpu sorted"?

Thanks

John

[Sugam Rangan]

Hi,

When you face the drop in the ping, is it the symmetric pattern for those 8 drops out of 100 or the continuous 8 drops? 

[riazay001]

Dear sugamrangan

Sometimes it is symmetric pattern and sometimes it is continuous. When I telnet the switch, I can feel like the slowness as I type a command and it is shown after a second and sometimes very smoothly.

[Sugam Rangan]

Probably there must be some process running which is utilizing the switch resource at max for few few seconds. Can you paste the full output of "sh process cpu history" and "sh process cpu" here.  

[Sugam Rangan]

Also, keep two sessions open. One from where you are executing the continuous ping and second for the switch you are facing issue.

As soon as you start observing the packet loss please execute the command "sh pro cpu his", you may see any single spike to 90% to 100%. If that's not the case then please check the interface errors on the trunk link between both switches on both sides. Because that's only the link from which you access the switch then that can be a same issue while telneting the device as well. 

[Gaurav Mahajan]

Do you have any time pattern when issue occurs ?

It could be due to some backup is running in the network during that time may result in latency in the internal network.