ESXi 8.0 on ARM!

The latest update to the ESXi-ARM Fling, which is now based on ESXi 8.0 Update 3b, is now available for download and installation. The Fling introduces an experimental version of VMware’s hypervisor tailored for ARM-based platforms, such as Raspberry Pi. However, note that it is not limited to ARM but could potentially be run on other ARM-based systems.

This release includes several new features and enhancements, such as support for vSphere Distributed Services Engine (Project Monterey) on ARM and improvements in stability and compatibility.

So, this is fun simply from an enthusiast perspective, but does it have real-world applications? Well, yes, one could argue this could be used in use cases like edge computing and remote office/branch office (ROBO) deployments where the workloads on top of ESXi aren’t too demanding. A particularly exciting use case could be using Raspberry Pi devices as an inexpensive vSAN Witness, enabling advanced storage configurations in a cost-effective way.

Installing it – as usual, requires jumping through a few hoops, but all in all it’s not that challenging to get started. Please refer to the installation instructions here:

https://higherlogicdownload.s3.amazonaws.com/BROADCOM/092f2b51-ca4c-4dca-abc0-070f25ade760/UploadedImages/Flings_Content/VMware-ESXi-Arm-Fling-PDFs.zip

I’ve opted to install this on my Raspberry Pi 4 (8GB) installed in an old Sharp MZ-2000 case. IT has dual monitors, but only micro-HDMI 0 is used. The other shows the standard stretched four pixels.

Using ChatGPT as a translation service

This is a quick note to show a concept of using the ChatGPT API for translation. In this case for English to Japanese, but it could be changed into any language. Since we want to talk to it we use Amazon Transcribe to turn speech into text. That text is sent to ChatGPT for translation. When the translated text comes back we feed it into Amazon Polly to turn it into speech again.

This is all running on a Raspberry Pi 3b+

Source code

Source code can be found on GitHub:

https://github.com/jonas-werner/chatBot

Quick demo video

Snakeinator case and final touches

The environmental control system for Mr. Snuggles the ball python is now nearly finished. Throughout the last couple of weeks I’ve worked on learning Fusion 360 in order to create a new case. The previous one was done with Tinkercad and the one before that in FreeCAD. The first iteration was a Tupperware case, lol

Work in progress while designing the case

The electronics didn’t end up as clean as I was hoping it would be but it works and that’s always been rule number one for this project. There will likely be a version after this and for that it would be great to etch a proper circuit board rather than using universal PCB boards as I have done up to now.

An old Wii U gave up its case fan for this project. It’s very quiet and perfect for case cooling

Currently I’m tuning the settings to dial in the right temperatures and humidity in the different parts of the enclosure. Grafana is a lifesaver as always 🙂

Work in progress on dialing in correct temperatures and humidity

The next step may be to add a humidifier of some sort and hook it up to the humidity values. The biggest challenge in controlling the environment has been the humidity. This seems to be a common gripe among those who use glass terrariums and I can certainly see why. Right now a couple of towels on top of the mesh screen in the lid help keep things under control but it’s not an ideal solution as it also limits air flow.

Mr. Snuggles loves his tubes