Cfengine contains many features and facilities that make it a powerful tool for system administration, but it has a large manual that is difficult to absorb without training. In this tutorial we assume that attendees have a basic understanding of how cfengine works and would like to develop a number of "best practices" and examples to maximize their returns.

Topics include:

• Review of some basics
• Automating deployment of software throughout your infrastructure
• UNIX/Mac/Windows
• update.conf
• cron and cfexecd
• When to run
• Integrating data from information sources
• Structure and organization of config
• The overlapping-set model
• Import
• Modules
• Methods
• When to use these tools
• Special functions and variables
• Variables, scalars, arrays
• Associative arrays and their limitations
• ExecResult, ReturnsZero, etc.
• ReadArray, ReadList, etc.
• IsNewerThan, IsDir, etc.
• Searching, matching, and wildcards
• Search filters
• Regular expressions
• Wildcard expansions
• How does cfagent evaluate things?
• Thinking declaratively
• Ordering: When does it matter?
• Locks; What are they, and why are they there?
• Iteration over lists
• Control, actionsequence, alerts
• Services and security
• PP keys and exchange (trust model)
• Authentication stages
• Rule orderings
• IPv6 issues
• Peer-to-peer services
• Example: Backing up laptops
• Host monitoring
• cfenvd
• Interfacing to tcpdump
• Understanding cfenvgraph output
• PeerCheck neighborhood watch
• FriendStatus function
• Future developments and discussion

Who should attend:
System administrators with a working knowledge of cfengine (or who have attended the introductory course) and who wish to extend their understanding of cfengine with examples and usage patterns. UNIX and Mac OS X administrators will be most at home in this tutorial, but cfengine can also be used on Windows 2000 and above.