These Get Started tutorials walk you through the basics of getting SaltStack up and running.
Every management tool and their Git forks claim to be flexible, but when we say flexible at SaltStack, we mean entire-management-approach flexible.
SaltStack can be implemented to follow the most popular systems management models:
The standard model for configuration management. A management server sends commands and configurations to agents running on large numbers of systems, and results are returned to the server.
If you have avoided management tools because you don’t want to set up all of that “extra stuff”, this is likely the model for you. Salt agents can perform management tasks with or without the involvement of a server.
The best thing about this model is that it is already built-in; if you don’t want to use a management server, you simply don’t set one up. You can move back and forth between this model and the Agent and Server model, or use them both together.
Salt commands can be executed on remote systems using SSH instead of the Salt agent. Configure all of your Web servers using the Salt agent, but perform ad hoc management tasks on your database servers over SSH. It is completely up to you.
The initial sections of this Get Started Guide demonstrate SaltStack using the Agent and Server model to keep things simple. If you are planning on using a different approach though, don’t let this get in your way. The main difference between each management model is the SaltStack executable that you use and how it is called. When it comes to doing work, SaltStack capabilities are fundamentally the same across each model.
SaltStack supports both of the leading methodologies to define system configurations. Follow a declarative methodology using SaltStack’s powerful requisites system, or go imperative with SaltStacks built-in ordered execution.