02-18-2019 04:42 PM - edited 03-08-2019 05:22 PM
Hi all
C3850# sho processes cpu sorted
CPU utilization for five seconds: 95%/39%; one minute: 95%; five minutes: 94%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
140 4086466 627425 6513 43.00% 47.00% 46.65% 0 NGWC L2M
233 87104868 126847578 686 7.15% 7.40% 7.34% 0 Spanning Tree
117 28933447 581712813 49 1.08% 1.02% 0.98% 0 IOSXE-RP Punt Se
what cause it ?
Thanks for any answer
Thank you
Solved! Go to Solution.
02-18-2019 07:22 PM
Is it the version 16.06.05 will be stable ? or 16.3.7?
For testing, we have upgraded a couple of switches to 16.3.7 recently. These switches seem to be working fine with no CPU issues so far but as I said before they all have bugs. I completely understand the 24x7 operation but in some cases like this you have no choice but have a maintenance window during the least busy time and upgrade. The other option would be not to upgrade but you constantly have to worry about the switch's CPU reaching 100% and crashing, which will cause more downtime for your environment.
HTH
02-18-2019 04:52 PM - edited 02-18-2019 04:56 PM
Hi,
High CPU is usually because there is a bug in the IOS you are running. Upgrade to a different version and test again.
The current suggested version by Cisco is:
cat3k_caa-universalk9.16.06.05.SPA.bin |
HTH
02-18-2019 05:35 PM
02-18-2019 05:59 PM
Oh dear , Because of too many bug, I upgraded the version from 3.7.4E to 16.3.6 just few days ago.....
I completely understand. Unfortunately having bugs in Cisco IOS is a fact of life. Now, some versions have more bug than others.
You just have to test to see which works for you best. Since you are running 16.x have a look at this document.
HTH
02-18-2019 06:09 PM
02-18-2019 06:24 PM
Are you running IPv6? If yes, go ahead and enable it and see if the CPU number for
NGWC
goes down.
HTH
02-18-2019 06:34 PM
02-18-2019 06:42 PM
If the CPU is normal now, no need to enable it. I would say for testing, enable it when the CPU is running high to see if this makes any difference. At the end of the day, even if this fixes the issue temporarily, you would have to upgrade the software to fix the issue permanently.
HTH
02-18-2019 07:04 PM - edited 02-18-2019 07:05 PM
Thank you for your recommend
upgrade the software is not a simple thing , customer use the network for 24 hours
what I concerned is that When I upgrade to a new version another bug we encounter, just like a few days ago what I did
Just as you say , I think it is necessary to upgrade, Is it the version 16.06.05 will be stable ? or 16.3.7?
Thank you
02-18-2019 07:22 PM
Is it the version 16.06.05 will be stable ? or 16.3.7?
For testing, we have upgraded a couple of switches to 16.3.7 recently. These switches seem to be working fine with no CPU issues so far but as I said before they all have bugs. I completely understand the 24x7 operation but in some cases like this you have no choice but have a maintenance window during the least busy time and upgrade. The other option would be not to upgrade but you constantly have to worry about the switch's CPU reaching 100% and crashing, which will cause more downtime for your environment.
HTH
02-18-2019 08:13 PM
02-18-2019 08:35 PM
NGWC = Next Generation Wireless Controller.
For 3650/3850, Cisco no longer supports NGWC after 16.3.X firmware.
04-09-2019 08:59 AM
Hi,
I have a similar issue with CPU using Denali 16.3.6... your upgrade to 16.3.7 solved this issue?
Regards,
Juan
10-23-2019 07:25 PM
Hi Juan,
Did you solved this issue?and how to solved this issue?
BR
Andy
10-23-2019 07:26 PM
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: