CML v1.1 simulation locks up entire VM

[Seb Rupik]

Hi there,

I've just finished upgrading to v1.1 excited to use the new IOSvL2 image, however stability of the entire platform seems to have been compromised. Whenever I start a new simulation, and then attmept to stop it, I lose contact with the VM, connections timeout and ultimately I have to reset the VM.

Before the simulation starts....

uwadmin looks happy enough:

VM CPU usage, flat  and low enough:

Simulation started, with a timer (in case something goes wrong and I lose contact!):

Now I try stopping the simulation, and this is where it gets stuck. Progress bars remain stationary:

uwadmin, shows timer not decrementing since started, more importantly page won't referesh:

All my CPUs consumed, unable to even use the vcenter console to interact with the VM:

At this point the VM needs to be reset.

What I'd like to know is if there is some way to interact with the VM other than the CML-eclispe client or the uwadmin GUI to shut the simulation down.

Then I'd like to know why is it performing so badly when the simulation does not exceed the VM capabilites?

cheers,

Seb.

[rkochery]

Hi Rupik,

May i know the resources assigned to CML, Please try to increase the CPU and RAM and see how it goes. 

[Seb Rupik]

Hi there,

It is currently configured with 4 vCPUs and 8GB RAM.

I realise 16GB RAM is the recommendation, but currently none of the monitoring is showing it as in contention.

cheers,

Seb.

[rkochery]

Are you able to run simulations succesfully??

[rkochery]

Hi Seb,

 If you are CML Corporate edition customer, please log a case Cisco CML TAC Support to further look into this issue.

Thanks,

Romal

[Seb Rupik]

Hello again,

I tried again this morning running the same 6 node simulation.

See attached vmware CPU graph. The fist spike is using 4 vCPU, second is 6 vCPUs and the last steadily graph is using 10 vCPUs.

uwmadmin page eventually became unresponsive, but does show simulation was within capability of the VM:

cheers,

Seb.