solo.rb

chef-solo configuration files will cut down on the amount of command line arguments you need to pass to chef-solo. With the following configuration, you could just issue chef-solo and you’re off to the races. By default, chef-solo will look for a solo.rb file at /etc/chef/solo.rb. To override this pass it the -c <path> argument, for example:

chef-solo -c /etc/chef/solo.rb -j /etc/chef/dna.json

For more gory details or to further tweak your defaults, check out the Chef Configuration Settings wiki page.

As quoted from the reference: “Where to locally store cache files like cookbooks and other transient data”. So uh, yeah… just set a temporary path here and move along.

The location chef-solo goes looking for chef cookbooks. This could also be a list if there are multiple directories. A list, however does not appear to be used much in the wild

The path to your JSON attributes. Also widely known as your Chef DNA. There’s where you are customizing want you want configured and tweak attribute settings (like listing users, packages, versions, etc.).

The default log level for a chef-solo execution. :info will give you a nice resource-leve description of what is going on (i.e. “Creating file[/etc/cool/mojo.cnf] at /etc/cool/mojo.cnf” or “Setting mode to 644 for file[/etc/cool/mojo.cnf]”).

For development and troubleshooting, I’d recommend getting used to the :debug level output. You’ll see exactly what commands get executed, if it’s skipping resources, the order it loads cookbooks, etc.

To increase the verbosity of logging for a one-off execution, pass chef-solo -l debug as an argument.

You can redirect the log output to a file or STDOUT. STDOUT may not be the ideal in production as you might want to have a log of what happened when (this can be very useful for audits or blames).