Here’s what I know so far:
Puppet is a tool for managing the configuration of large sets of hosts. It can be used for software provisioning, enforcing configuration, automating repetitive tasks, user management, etc.
Puppetlabs tries to describe its process as “Define, Simulate, Enforce, Report”: PL_howpuppetworks_notitle.png
Each managed system has a Puppet agent installed, reporting to the Puppet Master, this reporting is referred to as ‘facts’. The Puppet master compiles the facts into the ‘catalog’. This ‘catalog’ data is compiled and returned to the agent about what state it should be in. Any required changes are made to bring the system into the required state (in ‘no-op’ mode it just simulates the actions). Then a ‘report’ is sent back to the master. This report is available via open API for integration into other systems (CMDB I guess? Also monitoring solutions, Xymon, Nagios etc.).
Next step: grab the learning VM: http://docs.puppetlabs.com/learning/ The Learning Puppet series
No comments:
Post a Comment