Packer VS Vagrant

To manage a VM, you can use something as simple as just manual actions over SSH, or can use tools like Ansible, Hashicorp's Packer and Terraform or other automations. For an app where there is minimal load and security/reliability concern, VMs are still a great option that provide a lot of value for the buck

Server templating: Using Packer has never been easier to create reusable server configurations in a platform-independent and documented manner.

I'll give you hard, but rewarding and easy to modify(once you know what you're doing) way. Packer may be a thing you're looking for.

It is straightforward to build them with packer (I have built AMIs for 13.0 and 13.1, but 13.2 should be exactly the same). I've been meaning to write a blog post about it for a while, but have not gotten to it yet... In any case, what I am doing is using the EBS Surrogate Builder to start an instance running the official FreeBSD 13.2 image with an extra volume attached and run a script to create a zpool on the extra volume and bootstrap and configure FreeBSD 13.2-RELEASE on it. After that packer takes care of creating an AMI out of that extra volume, so you can use it... If you have any issues, let me know, and maybe I will finally get to writing that blog post...

HashiCorp Packer is a tool for creating machine images for a variety of platforms, including AWS, Azure, and VMware. It allows you to define machine images as code and supports a wide range of configuration options.

Packer is a tool that can be used to build machine images. Basically, it takes a base image, runs a series of steps to provision that image, and then burns a new image. In my workplace we use it heavily to build AWS AMIs. But it has an ARM plugin that looks to be very very suitable for building customised Raspberry Pi images (my quick read of the doco there says it can go ahead and write the final image to an SD card for you too).

Seems related to https://github.com/hashicorp/packer/issues/12118 But the workaround discribed in the comments doesn’t seems to work anymore

I'd suggest using a tool like Packer to build a gold image based on your base AMI and all your changes. Then you'll have your own AMI you can launch new instances with.

Your assumption sounds correct. It sounds like you shouldn't focus on migrating specific instances, but instead migrating the template image used for autoscaling into GCP. I tend to prefer Packer for this job, or otherwise recreating the golden image directly on GCP.

Take a look at Vagrant! https://www.vagrantup.com/ In my admittedly limited understanding I believe it offers closer to a nix like reproducable rather than repeatable deployments.

on the off chance one hasn't been tracking it, there were several "we don't need your stinking BuSL" projects when this drama first started:

https://github.com/opentofu#why-opentofu (Terraform)

https://github.com/openbao/openbao#readme (Vault)

and I know of several attempts at Vagrant <https://github.com/hashicorp/vagrant/forks> but I don't believe one of them has caught traction yet

There are also some who have talked about an "open Nomad" but since I don't play in that space I can't speak to it

It sounds like you're using Nix as a sort of configuration management solution. CM just isn't worth it for managing a single desktop IMO. It triples the effort for whenever you need to add or remove a package, as you must now add that also to your nix configuration. You're supposed to be able to make that back up in time saved restoring to the next machine, but inevitably the next machine will be different enough that you'll have to edit it all anyway. In the end I just got tired of trying to manage my own machine with infrastructure as code (though in fairness I was using puppet at the time not nix).

I keep a git repository with all my dot files in it[1]. This seems to work the best. It has a Windows folder as well, and I copy that out whenever I need to set up Windows.

A lot of people like using WSL but I hate how it hogs on my memory. Hyper-V is a terrible virtualization engine for consumer-grade use cases because it can't thin provision RAM. If I need to use docker, I will spin up a small Linux VM using vagrant[3] with Virtualbox[4] and put Docker on there. Vagrant is an extremely underrated tool in my opinion, particularly in a Windows context.

I use scoop for packages. Typically I will scoop install msys2 and then pin it so that it doesn't get blown away by the next upgrade.

Then I basically do all of my development inside of msys2. I can get most things running in there without virtualization. In my case that means sbcl and roswell for common lisp, senpai for irc, and tmux and nvim for sanity. Msys2 uses the pacman package manager and this is good enough.

All In all, I set up my Windows machine affresh after a while of not using it and it took me about 3 hours. Most of that time was just getting through upgrades though, I felt like it was pretty fast.

1: https://git.sr.ht/~skin/dotfiles

2: https://www.msys2.org/

3: https://www.vagrantup.com/

4: https://www.virtualbox.org/

Tools like Docker and Vagrant can be used to allow local environments to mimic production environments.

There's an open issue [1]. A scripting interface has since been added [2], and updated [3], so there's progress.

[1] https://github.com/hashicorp/vagrant/issues/12518

Someone should fork and maintain Vagrant with an MPL open source license:

https://github.com/hashicorp/vagrant

https://github.com/hashicorp/vagrant/blob/v2.3.7/CHANGELOG.m... ?

The changelog lists both improvements and bug fixes and there's even apparently some effort to port it away from ruby: https://github.com/hashicorp/vagrant/blob/v2.3.7/internal/cl...

If you’ve ever encountered the dreaded “Vagrant fatal error: runtime BSDThread_Register error,” you’re not alone. This perplexing error message can be frustrating and confusing, especially if you’re new to Vagrant and virtualization. But fear not! In this article, we’ll unravel the mystery behind this error, explain its meaning, and provide solutions to help you overcome it.