What's the best way to include the following steps in my deployment: • Once an EC2 instance is up, copy a shell script over to it over SSH • After the copy is complete, run the shell script on the EC2 instance. Is it https://github.com/pulumi/pulumi-command ?

That's what I do. I first have some code that checks the instance state in a loop until it returns "running". In javascript, you'd get the ec2 instance via:

let instance = await aws.ec2.getInstance({ instanceId: id });

And then look at

instance.instanceState

. After that equals "running", I use a pulumi remote

Command

that executes

timeout 15m cloud-init status --wait

to ensure that the instance is fully booted and done with running any init scripts. Then any later commands/file copies simply have a dependency on the

Command

that waits for cloud-init to be done.

Thank you!

or you could give a cloud-init script when you create the instance cloud-init is included in the major cloud distributions, and its purpose is exactly that : run a script on the first boot