I've had positive experience with https://github.com/PowerShell/openssh-portable on Windows Server 2016
- I've been able to install very simply via chocolatey (https://github.com/PowerShell/Win32-OpenSSH/wiki/Install-Win32-OpenSSH) (there's a non-choco method that I have not tried out, since I expect you might not want to take another dependency to a base image)
- I've been able to successfully connect over ssh from my macOS to the Windows machines
- I'm successfully using the
ssh-agent service to hold private keys for user accounts
I would love a world where I didn't have to deal with WinRM (can elaborate, but on request - want to keep this concise) and instead could use ssh to communicate to each of my remote fleet.
I was wondering at which point GCE might start to create base images that have it installed and, if not running, at least ready to go via a metadata-script nudge? I appreciate that one could install it via those means too, at the cost of extra time during either baking or boot.
Would you welcome a PR that added it, and if so, would you mind outlining what you might, vs might not, accept?
I've had positive experience with https://github.com/PowerShell/openssh-portable on Windows Server 2016
ssh-agentservice to hold private keys for user accountsI would love a world where I didn't have to deal with WinRM (can elaborate, but on request - want to keep this concise) and instead could use ssh to communicate to each of my remote fleet.
I was wondering at which point GCE might start to create base images that have it installed and, if not running, at least ready to go via a metadata-script nudge? I appreciate that one could install it via those means too, at the cost of extra time during either baking or boot.
Would you welcome a PR that added it, and if so, would you mind outlining what you might, vs might not, accept?