CML NX-OS 9000 won't start - stuck in QUEUEING. Console will not open.

[noisey_uk]

Cross-post from learningnetwork.cisco.com

Running on a Dell R610 with 2 x Intel(R) Xeon(R) CPU X5680 and 60GB RAM. ESXi 6.7.0. Build 17167734.

Behaviour seems to be the same for CML 2.1.0 and CML 2.1.1.

Have reinstalled multiple times. Even tried installing updates from Cockpit although there's a pinned advisory against doing so - no improvement.

refplat ISO is mounted.

I've seen a few posts which suggest this is happening to others too.

Any ideas? Simulating NX-OS for study and labbing is the only reason I bought this product. I was hoping to take an exam next week! The NX-OSv Titanium seems to do weird things with CAM tables that means things like OTV don't work properly so I think I need the 9000.

[balaji.bandi]

*Must support VTx and EPT or AMDv and RVI. These CPU flags are required for nested virtualization.

Look at the system requirement :

https://developer.cisco.com/docs/modeling-labs/#!system-requirements

https://developer.cisco.com/docs/modeling-labs/#!installing-cml-2-1

[noisey_uk]

I believe the Intel(R) Xeon(R) CPU X5680 supports both of these features. https://ark.intel.com/content/www/us/en/ark/products/47916/intel-xeon-processor-x5680-12m-cache-3-33-ghz-6-40-gt-s-intel-qpi.html

Also, wouldn't that affect ALL guest devices, not just the NX-OS 9000? e.g. the unmanaged switch, Alpine, NX-OSv Titanium run fine... surely they wouldn't if VTx and EPT were the issues?

[balaji.bandi]

Rest all working , then your nested settings ok.

i would not go to titanium, i will more focus on nexus image of 9K  - i run the lab with nexus 9K ok with 2.X CML.

[noisey_uk]

OK... so if the nested settings (VTx and EPT) aren't the issue, what else might be? Any idea which log files to check etc.?

[balaji.bandi]

i look for compute, what is the resource given to CML, i am sure nexus need good resources to start.

[noisey_uk]

CML is the only VM running. Resources are as per the deployed OVA. Did you increase the resources since deploying the OVA?

[balaji.bandi]

what is the resource please send the screenshot or increase CPU and RAM if you have more.

[noisey_uk]

The resources were 4vCPU & 8GB RAM - as per the original OVA. The CPU and memory status bar at the bottom of the CML workbench were always green, although the NX-OS 9000 would not boot (stayed QUEUED). I increased to 12vCPU and 32GB RAM as you suggested and it's now booting! Thanks @balaji.bandi!

Do you know of any documentation/guidance for the appropriate resources to give to CML? The minimum requirements clearly aren't accurate. It would be useful if CML gave an error rather than just sit there... or at the very least not indicate CPU and memory to be "green".

[noisey_uk]

Changed allocation to 8vCPU and 16GB RAM and NX-OS 9000 nodes are still starting.

[balaji.bandi]

Honestly some time cisco documentation is very huge, and some time not accurate, I do not believe with minimal CPU and 8GB ram you can not do vPC Labs - its all experience and reading some blogs and requirment of the image in cisco document.

https://developer.cisco.com/docs/modeling-labs/#!faq/reference-platform-and-images-questions

Glad all well. can we mark as resolved if no further assitance required.