adding a new lab machine after re-importing a lab is not possible

[carfesh]

This is not necessarily an issue per se, but probably more of a feature request.

It would be great if an existing lab could be extended later by running Import-Lab and then adding machine definitions to it, installing those explicitly.

Attempting to do so currently fails with an error message: "Create a new lab first using 'New-LabDefinition' before adding machines".

This would allow creation of more complex multi-part labs.

[raandree]

@PerPedersenMSFT has taken care of this. Recycling one of his statements:

Simply add the machine to the script and re-run the entire part of the script which creates the machines.
Ie, add a line with Add-LabMachineDefinition -Name NewMachineName and run the lines beginning with New-LabDefinition ending with Install-Lab line(s)

I am not sure if this works with all roles but definitely with AD.

[carfesh]

Thanks for the idea, I will give it a shot! For the lab I'm working on this might become complicated however because the lab can be installed in different ways (the subnet can be customized, machine OS can be customized, it can be installed with or without an externally facing adapter, etc.)

I will need to find a way to keep those settings "in sync" so they don't contradict each other, the easiest way would really have been to just re-import the existing lab definition. Other things like installing software etc. also work fine when doing this, but sadly not the creation of additional machines.

Are you aware of any technical or design restrictions why this isn't possible? Maybe I'll look into that myself when I have a little more time to spare..

[BristleBeard]

Perhaps when a lab is created, a 'state' file could be defined with it, basically marking lab creation steps as done or complete as they execute.

For example, you create a lab with a DC, a root CA, and two member servers. Once it is finished, the basic steps involved (create domain, join domain, create each server, join each server, etc) could be marked as completed in an xml file. If you add a new definition to the lab later, when you execute the script, it checks against the xml and finds that all steps are completed except there is no info for this newly added definition, so it does the steps needed to only the newly defined server.

[Juvls]

Yes it would be great to have this feature implemented!

[raandree]

Closing this. We continue to track this on #123, which is about almost the same.