High Memory Utilization:2960G

[jpl861]

Hey guys,

I just want to ask if you are experiencing the same thing on your WS-C2960G-24TC-L switch. According to show version, the DRAM of the switch is 64MB but when I do show memory free, the total is only 25MB. When I checked our monitoring server, it can only see 25MB as well. According to the output of show memory debug leak, one process is leaking.

sh memory debug leaks

!Output Cut

                 Processor memory                

Address    Size   Alloc_pc  PID  Alloc-Proc       Name

28A5A14       80   5E3774  0    *Dead*          hulc_flash init

We experienced this on 12.2(58)SE1, then we upgraded to 12.2(58)SE2, then finally to 15.0(1)SE. There's an open case with Cisco already and now they are claiming that high memory should be normal behavior. Even after power cycle with no desktops on it, the utilization of the memory is 71%. We have five 2960G switches running 12.2(58)SE1 and they are all running 85% memory and sometimes, the switch cannot serve the workstations as we are running 802.1x. The LEDs are amber and cannot authenticate the workstation because of a very high memory utilization. I cannot telnet nor console the switch because of no memory.

Cisco said that I shouldn't see all the 64MB from the CLI but 25 out of 64 is unacceptable I think.

I have another 2960 switch in another center which is WS-C2960-24PC-L, also 64MB but show memory and solarwinds can see 33MB but it is only eating 25% of memory.

Any idea why we have very low memory?

Thanks in advance.

John

[Somasundaram Jayaraman]

Hi John,

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

++ Have you tried reloading the box, because while reloading the memory will be defragmented.

++ I checked and i am not finding any bugs related to this.

Cheers
Somu

[jpl861]

Thanks for the reply.

The hanged process is hulc_flash init. I have already tried to reload the box and the issue is the same. 71% memory utilization even without workstations. Only uplink ports back to core.

[smogra]

Can you post the outputs of following:

1. sh mem summary

2. show memory allocating-process totals

3. sh mem dead

and sh log as well if you see any error mesages there related to memory/memory allocation failure.

Cheers

Sweta

[jpl861]

Hi Ma'am...I have attached SWI007.txt for the switch output. Thank you so much.

[rogeriomonteiro]

Hello guys,

I am with an Cisco 2960G and i did the upgrade for the IOS to the newest version 15.0(2)SE11, but i have a issue with your memory processor who it is high than 85%, i had just done the reboot but nothing changed, so who can help me?

[rogeriomonteiro]

Have you solved this problem already? I am with the same thing.

[rogeriomonteiro]

Have you solved this problem already?