High CPU - Hulk Led Process

Andy Roberts · ‎03-17-2011

Anyone seen this before.

Have just built a stack using five WS-C3750V2-48PS

The stack is up and running. Nothing connected so no spanning tree issues or flapping links. Any idea why the CPU is so high on what is an idle switch.

Culprit seems to be Hulk LED process

CPU utilization for five seconds: 23%/0%; one minute: 39%; five minutes: 29%

135 48785862 6314845 7725 15.97% 13.74% 14.77% 0 Hulc LED Process

IOS Version I am running is c3750-ipservicesk9-mz.122-50.SE1

Am worried that if it's like this now, once we start adding data and voice and actually switching it is going to go thorough the roof.

Thanks

Andy

schooram · ‎03-17-2011

Hi Andy,

In the older version of 3560/3750, this issue was reported and logged as a bug. 
Eventually the issue was resolved for the older version of switches. 
The bug id was CSCsi78581 and it was fixed in 12.2(44)SE. 

http://tools.cisco.com/Support/BugToolKit/search/getBugDetails.do?method=fetchBugDetails&b
ugId=CSCsi78581

In newer versions of 3750 switches, same issue was reported but this time the
developers confirmed that the process utilization is by design. As documented in bug CSCtn42790

Hulc LED Process - Is responsible for link status detection.

Symptom:

Hulc LED Process uses 15-30% CPU on Catalyst 3560X/3750X platforms.

Conditions:

The is seen in 12.2(53)SE releases or later.

Workaround:

This is an expected behavior and there is no workaround.

Further Problem Description:

The "Hulc LED" process does following tasks:

- Check Link status on every port
- If the switch supports POE, it checks to see if there is a Power Device (PD) detected
- Check the status of the transceiver
- Update Fan status
- Set Main LED and ports LEDs
- Update both Power Supplies and RPS
- Check on system temperature status

Bug details at :

http://tools.cisco.com/Support/BugToolKit/search/getBugDetails.do?method=fetchBugDetails&bugId=CSCtn4279

hence, this is expected behavior and you should not worry about it.

Best Regards,
Sandeep

Rob Vessey · ‎04-14-2011

Thats all well and good, but the Hulc LED Process on ONE of my 3560X units is

constantly at 60+%!

NOC-CS35X-RBZ1-24U42#sho proc cpu sort
CPU utilization for five seconds: 93%/0%; one minute: 94%; five minutes: 94%
PID Runtime(ms)   Invoked      uSecs   5Sec   1Min   5Min TTY Process
138   7545509391049206947        719 66.87% 61.50% 62.88%   0 Hulc LED Process
47   116687231 359726790        324 15.67% 15.69% 15.79%   0 SFF8472
192   166031836 166979615        994 1.59% 1.62% 1.63%   0 Spanning Tree
102   714876232 54005931      13237 1.43% 1.56% 1.59%   0 hpm counter proc
239   8187570861830165431        447 0.79% 0.24% 0.19%   0 REP LSL Hello PP
   4   137749233 10987536      12536 0.47% 0.39% 0.33%   0 Check heaps
319        3163       868       3644 0.47% 0.22% 0.16%   2 SSH Process
147   156021790 12912747      12082 0.31% 0.33% 0.33%   0 HQM Stack Proces
37    35251112 22562070       1562 0.31% 0.13% 0.11%   0 Per-Second Jobs

The console response is terrible. What might the problem be here? I've tried to find a workaround or fix, but have not had any luck at all.

We aree running the 12.2(53)SE2 universal image with ipbase only.

tj.mitchell · ‎04-14-2011

The listed bug is correct, the LED process will hover like that. It's the expected behaviour of the and there's not a thing you can do. Least that is what TAC told me.

According to whatyou have posted though, there are other issues. STP process is fairly high, it shouldn't be that high. Once STP is completed, then the process should stay much lower than that - like .0something% not 1.x something%.

That switch is getting pushed because of something and I don't think it's this process.