I have been working on command line deployments of aws and azure, that include the network and application for a full app stack test. It is very rudimentary, but what I have discovered it, its an easy way to edit some variables, then launch configurations, networks, objects, gateways, apps, etc, in varying combinations to see first hand what works, and what doesn't. Not sure if you have network connectivity between two VPC or subnets? Don't wonder, spin them up and test for yourself. I even have simulated clients in the script, but be careful with them, they can get aggressive if launched from a decent sized ami. They will consume the web app at a ferocious rate. 4 or 5 clients running on 16 core machines can consume over 100,000 concurrent connections, and sustain 10k+ connections/sec setup rate. If you aren't seeing that high a transaction rate, try running up the application as well as the gateway, or spread the traffic across multiple sites. I don't have elb configuration in this, yet, but DNS is configured, and if you have played with DNS in aws, it's not really DNS, its a programmatic change to the enviroment. My hosted value for DNS is in there now, not that you can do anything with it, you will need to set that to use DNS injection, otherwise ignore the dns errors, it just means there is no injection for name to IP. Everything will still work based on IP.
Azure is also incorporated, but the app needs to be manually installed, just time on my part. I will be putting together real training and documentation around using this, but for now, wanted to share this as it is live. For those of you that tried and say cpdeploy, the shell script, here is the container version, environment all set with the tools you need. Edit /cpdeploy/VAR, or type 'cpconfig' to set variables. To load the latest container on any docker host capable of running docker containers;
docker run --name cpdeploy -i -t kellman/cpdeploy
that will pull the latest container and drop you in a shell to try for yourself. Your milage will vary. This will continue to evolve and expand, and I will use this location to update the progress. With this container, I am killing all work on the original shell script. It is still online, but will not be updated. Please use the container for future versions. Big thanks to Lon for his curl based R80api calls, makes this much easier to transport, nice work Lon Kaut .