I tested these last week, and they got me about 90% of the way on two custom test containers (Forgejo and Postgres). I especially like the fact that they are a lot like microVMS instead of pure containers. Given that these are pure Windows services, it eases a lot of non-technical senior managers minds in my highly regulated environment.
That said, this is just another example of Microsoft's extend and extinguish past behavior.
it's a linux container but through WSL and natively integrated into windows. that's basically it.
It replaces the need to install docker and mess with a lot of hackery instead with an officially first party supported linux container subsystem.
Edit: and since I forgot to mention a key value-add of this is that it's integrated with group policy, defender etc so you can limit specific users or groups or OUs, etc to running specific containers or to exposing specific resources/network devices/hardware/etc to WSL at all. It makes managing all of it no different than managing any other part of an enterprise windows setup.
And likewise WSL has native antivirus/anti-intrusion instrumentation hooks via a Microsoft Defender for Endpoint/MDE plugin to WSL so that unlike with docker, etc you don't get a sudden blindspot that's non-trivial to monitor (or that comes with a heavy perf penalty).
So it's not something for me, but like everything Microsoft makes, it's for our BOFH's who will torment us with yet another corporate compliance machine.
Windows Subsystem for Linux (WSL) is a feature of Windows that allows you to run a Linux environment on your Windows machine, without the need for a separate virtual machine or dual booting -- https://learn.microsoft.com/en-us/windows/wsl/about
Those who can’t or won’t move off of Windows have some other way to run Linux stuff in a virtual machine that integrates into a native terminal. I’ve found it quite awkward in general, and honestly haven’t thought much of it since I just decided to move to Linux.
As a Linux developer on Windows, I find this feature really compelling. I had been doing this by importing debian and alpine images manually, to reduce the scope of agents to a limited vm instance with limited credentials.
With WSL, you can control which Windows dirs are accessible (read or read-write). you can easily clone new VMs (distros ) as a new instance.
WSL containers adds convenience and consistency to the manual steps I was doing before.
for example: a primary debian developer env, a second Debian maintainer/developer instance, multiple alpine containers (for each agent stack, one with cloud keys, one with messaging keys).
This adds a very feature rich vm and container experience on top of WSL2 , which was great to begin with.
In case anyone was curious, this works like docker desktop on Windows.
It launches a Linux vm which then runs the container(s).
is it compatible with docker cli ? can i stop running docker desktop? mine hasn't been the most reliable
I tested these last week, and they got me about 90% of the way on two custom test containers (Forgejo and Postgres). I especially like the fact that they are a lot like microVMS instead of pure containers. Given that these are pure Windows services, it eases a lot of non-technical senior managers minds in my highly regulated environment.
That said, this is just another example of Microsoft's extend and extinguish past behavior.
Can someone explain what the use case for this is?
it's a linux container but through WSL and natively integrated into windows. that's basically it.
It replaces the need to install docker and mess with a lot of hackery instead with an officially first party supported linux container subsystem.
Edit: and since I forgot to mention a key value-add of this is that it's integrated with group policy, defender etc so you can limit specific users or groups or OUs, etc to running specific containers or to exposing specific resources/network devices/hardware/etc to WSL at all. It makes managing all of it no different than managing any other part of an enterprise windows setup.
And likewise WSL has native antivirus/anti-intrusion instrumentation hooks via a Microsoft Defender for Endpoint/MDE plugin to WSL so that unlike with docker, etc you don't get a sudden blindspot that's non-trivial to monitor (or that comes with a heavy perf penalty).
So it's not something for me, but like everything Microsoft makes, it's for our BOFH's who will torment us with yet another corporate compliance machine.
Windows Subsystem for Linux (WSL) is a feature of Windows that allows you to run a Linux environment on your Windows machine, without the need for a separate virtual machine or dual booting -- https://learn.microsoft.com/en-us/windows/wsl/about
WSL is a VM, along with all the overhead and inefficiencies that come with it.
Those who can’t or won’t move off of Windows have some other way to run Linux stuff in a virtual machine that integrates into a native terminal. I’ve found it quite awkward in general, and honestly haven’t thought much of it since I just decided to move to Linux.
As a Linux developer on Windows, I find this feature really compelling. I had been doing this by importing debian and alpine images manually, to reduce the scope of agents to a limited vm instance with limited credentials.
With WSL, you can control which Windows dirs are accessible (read or read-write). you can easily clone new VMs (distros ) as a new instance.
WSL containers adds convenience and consistency to the manual steps I was doing before.
for example: a primary debian developer env, a second Debian maintainer/developer instance, multiple alpine containers (for each agent stack, one with cloud keys, one with messaging keys).
This adds a very feature rich vm and container experience on top of WSL2 , which was great to begin with.
NULL
[flagged]
I think a lot of people in enterprise environments do.