Root cause of high latency

  • 1
  • 1
  • Question
  • Updated 2 years ago
  • Answered
Hi guys,

I need your expertise to help me troubleshoot my issue. My current infrastructure have summit x450a(redundant) as core switch and blade servers(hosting VDI).

However, I noticed there is high cpu utilization on those core switches. Plus I can't ping the VDI vlan eventhough there is no acl. The VDI don't have any issue connecting to the internet.

I attached some screenshots for your reference. What I really want to know is where is the root cause of this issue. For your info my switches use XOS 12.XX

Thank you in advance


Photo of hafiz

hafiz

  • 196 Points 100 badge 2x thumb

Posted 2 years ago

  • 1
  • 1
Photo of Edward Tsui

Edward Tsui, Employee

  • 602 Points 500 badge 2x thumb
what is the uptime of your switch?
Photo of Ron Huygens

Ron Huygens, Employee

  • 2,878 Points 2k badge 2x thumb
Hi Hafiz,

A good start is to find out which process is causing the high CPU usage using the command "top"
https://gtacknowledge.extremenetworks.com/articles/How_To/Understanding-the-output-of-the-TOP-comman...

Usually, high CPU utilization needs to be diagnosed case by case.
If you are seeing network problems or problems on the switch you can check the "top" output.
If you see some processes each running at maybe 10 or 20% for a longer period, I would recommend opening a case with GTAC so it can looked into further.
Photo of hafiz

hafiz

  • 196 Points 100 badge 2x thumb
Here is the attachment after I ran top command. as u can see in the first row, is it possible to have high CPU utilization eventhough the STAT is "S"(sleeping)?
Photo of hafiz

hafiz

  • 196 Points 100 badge 2x thumb
just 1 week plus.because we have power trip last weekend.but this issue already happened almost 1 month
Photo of Jan Steinbach

Jan Steinbach

  • 1,078 Points 1k badge 2x thumb
You could check if there is a loop via ELRP command and in addition, use the "show port utilization" to look for Ports with unusual High packets-per-second.
Photo of bw447

bw447

  • 906 Points 500 badge 2x thumb
We had some latency issues a while back.
Command we used a lot was "show port statistics" It does a good job of showing RX/TX , Broadcast and Multicast packet counts.
We saw that we had a high amount of Multicast. GTAC recommended we upgrade our EXOS version. We are currently running 16.1.1.4. Since we did that we haven't had any issues come up.
Photo of hafiz

hafiz

  • 196 Points 100 badge 2x thumb
This is the result after I ran the command :)
Photo of hafiz

hafiz

  • 196 Points 100 badge 2x thumb
I think I have found the solution.but I haven't  try yet.refer below :)

https://gtacknowledge.extremenetworks.com/articles/Solution/High-CPU-Utilization-from-Process-Nodemgr