I've seen that many discussions already have been raised about the high CPU utilization of the Hulc LED process on 3560X-48PS. We face the same issue on all of our 3560X with Hulc LED process of around 20% to sometimes 40%. Displaying the process cpu history shows that it is constant around 40%.
We also face performance issues on the devices with the Hulc LED process around 40%. It is the only process with this high utilization.
I read different infos about the purpose of this process, reaching from checking layer 2 traffic to checking the ports (PoE etc.).
Can anyone finally explain:
1. The purpose of the Hulc LED process
2. Why the CPU utilization of 20% is a normal condition ? In my opinion this far to high
3. Can the process be influenced by changing the configs (I read in one post that the command "switchport trunk allowed vlan ...." on the uplinks solved the problem)
Any hints and tips are very welcome
we will upgrade coming Sunday to 15.0.2-SE9 (recommended)
Who recommended going to this version (or 15.X version)?
What was the reason to go to this version?
it is Cisco's recommended version (see attachment)
I am familiar what is Cisco's "recommendation" for the 3560G/X & 3750G/X, however, I don't agree with the recommendation.
For 3560G/E/X &/or 3750G/E/X I would recommend 12.2(55)SE-train due to stability.
many thanks for your reply; I fully agree with you; I'm also familiar with the "recommendations", but I have some of the 3560x running with 15.x without any problems.