This message was deleted.

Can you explain more about your use-case? Generally, if one has different groups of resources that are managed separately, each group would be in its own stack.

Since we have so many dependencies and currently there isn't a way to manage a run between multiple stacks we opted to have everything in a single stack and it's been working well. I want to be able to be as granular as I want to be. A couple of scenarios: 1. Update all services on all environments 2. Update service A on all environments 3. Update all users If there was a way to tag/label resources, like we can with Ansible, then this opens up lots of possibilities. We could still use

-t

, but target a group of resources instead of specifying each one individually.

And to clarify, you have cases where what is defined in the pulumi code is not 100% in synch with the actual state of the stack? In other words, after you modify the pulumi code, you don’t want to just do

pulumi up

and let pulumi figure out which resources need to be updated?

Correct, in the same way you'd use

-t

to update a single resource. This is for targeted stuff only, I'm very happy that

pulumi up

does the whole thing.

Got it. So, currently, there isn’t a tagging/grouping mechanism as you describe. I did a bit of looking and didn’t see your use-case described in any existing issues. So, I would suggest opening an issue here: https://github.com/pulumi/pulumi

Thank you, will do