We are currently using Chef Solo (12.14.x) and are enabling Chef Client/Server. What are your thoughts on continuing to run Chef Solo for specific use cases, like controlling a deployment through execution of a limited runlist and node attributes while Chef Client/Server is also configured? What gotchas might we expect in this multi configuration?
Here at my company we don’t use Solo but we do something similar by passing an override run-list “-o run-this-cookbook-now” to the chef client which then fetches everything from the chef-server. I also have a small bash script which takes some arguments, writes them to a JSON file and then it runs the chef-client with a specific recipe passing the JSON as attributes.
So I don’t really see how Solo could be more powerful than anything directly connecting to he chef server.