WilliamLam.com

VCF 9.1.1 - OCuLink External Graphics (eGPU) Dock with integrated PSU for GPU Passthrough with VKS

by // 2 Comments

For those who have been following my VCF Private AI Services (PAIS) exploration with deploying my own AI models and using them with Coding Agents running on VCF, I have found that these AI workloads can really push the thermals on my NVIDIA RTX 4000 running inside my Minisforum MS-A2, eventually causing the GPU to start throttling.

Earlier this year, I had shared a solution from a colleague who wanted to leverage a more capable GPU with vSphere Kubernetes Service (VKS) and was able to externalize the GPU using an OCuLink-based eGPU dock that is also compatible with the MS-A2.

I was initially going to use the same eGPU solution, which would have also required purchasing a separate PSU to power the GPU. While looking at my options, I came across another OCuLink eGPU dock (AG01) from Aoostar that had an integrated PSU (up to 800W), was plug-and-play ready, and certainly had a much cleaner aesthetic design.

[Read more...]

Quick Tip - Improved NVMe Tiering Device Health Monitoring in VCF 9.1

by // 1 Comment

Unfortunately, one of my consumer-grade NVMe SSD devices that I had configured for NVMe Tiering (Memory Tiering) in my VMware Cloud Foundation (VCF) 9.1.1 environment had recently degraded. While vCenter Server had raised an alarm stating “NVMe Memory Tiering device is not healthy.”, I had dismissed it since the alarm did not provide any additional details about the underlying issue.

In VCF 9.1, NVMe Tiering usage can be conveniently viewed inn the vSphere UI under the Memory tile for either a selected ESX host or vSphere Cluster, which summarizes both Tier 0 (DRAM) and Tier 1 (NVMe Tiering) usage.


When it comes to understanding the actual health of an NVMe device used for NVMe Tiering, I was not sure whether looking at the SMART data was still the recommended approach or if there was a better way?

[Read more...]

Quick Tip: The Fastest Way to Clear Partitions for ESX Reinstallation

by // 4 Comments

I often redeploy my VMware Cloud Foundation (VCF) lab environment for a variety of reasons, from testing customer scenarios to testing the next version of our software. Because I have fully automated the ESX installation and configuration process in my lab, reinstalling ESX is a trivial task that takes less than 5 minutes from power-on to a fully configured system ready for VCF.

When reinstalling ESX, especially on systems configured with NVMe Tiering, vSAN, and local VMFS storage, the real challenge is clearing the existing partitions so the system is ready for a fresh installation. Traditionally, this can involve several manual steps and, in some cases, booting from a Linux live CD to remove stubborn partitions.

Given how frequently I redeploy my environment, I have developed the following workflow that consistently cleans up the various storage configurations without requiring a separate live CD environment. As I was performing yet another ESX reinstallation while writing this post, I realized I had never documented my workflow, so I thought I would share it in case it helps others.

[Read more...]