Hulc LED Process % Significantly Lower Than on Other 3750x

[cer43tcent]

Hi All.  I have a few Layer 2 3750x, lanbase, (C3750E-UNIVERSALK9-M), Version 15.0(2)SE4, PoE,  switch stacks and single switches.  I've noticed on one switch stack (5 members) that the 5 min cpu util is 20% and top three processes w/ 5 min are

RedEarth Tx Mana 4.79%

RedEarth I2C dri 3.00%

Hulc LED Process 0.90%

On another switch stack (3 members) that the 5 min cpu util is 38% and the top three processes w/ 5 min are

Hulc LED Process 19.39%

RedEarth Tx Mana 6.65%

RedEarth I2C dr 1.95%

I've compared the show ip traffic, show controllers cpu-interface, and show platform tcam utilization, but no statistics is significantly greater than the other switch stack.  Moreover, other switch stacks and single switches on the LAN have the same cpu util average of 40%.  Any troubleshooting ideas on why the 5 member switch stack is at a cpu util of 20% and the other 38%.

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

I believe the Hulc LED process, is a background process, if if so, it would have very little impact on normal functioning, i.e. other than for curiosity, this CPU utilization difference isn't worth pursuing.

[Massimo Baschieri]

Sure, but most network management products start sending alarms if total cpu utilization exceeds a predefined threshold (usually 50%), then users start worrying and calling for maintenance.

It's a very annoying issue.

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

Indeed, great point about CPU usage monitoring/alarming!

[Massimo Baschieri]

No one knows how to solve it or at least attenuate it?

It seems not related to a particular ios release or a particular configuration, at least in my experience

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

I've thought I've read it can vary per IOS version.

My guess is, as (and if) it doesn't impact actual device forwarding performance, Cisco hasn't tried to really minimize it.  Again, though, your point about monitoring/alarming is great because typical monitoring wouldn't "understand" high background CPU usage isn't a real issue.

PS:

I've had WAN links stayed pegged at 100% for long periods due to bulk background traffic.  QoS can insure such traffic isn't adverse to other traffic, but this too will cause typical monitoring to "alarm".

[InayathUlla Sharieff]

Please go through my document:

https://supportforums.cisco.com/document/12298401/troubleshooting-high-cpu-3750

HTH

[limmingchun]

I found out the same issue at my environment too. All my switches are 5 members-stacked switch, but some of it always show Hulc LED Process around 20%, some of it aroud 1%.

Finally able to found out the different between all this switches.

if all the members are 48 port switch, then it will be around 20%,

if the member consist 24 port switch, then it will be around 1%.

I do not know how to solve it and what the reason that caused this, even cisco also cannot explain about this.

Just to share with you that the different I found. I also created a discussion at support forum too :

https://supportforums.cisco.com/discussion/12394456/csctn42790-3560x3750x-elevated-cpu-usage-due-hulc-led-process