Populate the CMDB
This page sums up the options you could consider for populating the CMDB.
Via the web interface
Some items can be populated by hand. It's a good method if you run a test environment or if your datacenter has a small size (< 40 machines for instance). Plus for some items (contacts, applications), it may be the unique way for gathering all these informations.
Currently the web interfaces allows edition of :
- servers, network devices, storage devices
- applications & application instances
- definition of databases
Via rake tasks
This is the method I currently use at work. My CMDB is populated through file transfers and custom rake tasks. Some of these tasks are versionned in the current repository. There are 2 global tasks that I run every 10 minutes on my production environment for loading new items from files and cleaning up old stuff :
rake import:all cleanup:all
This is the current implementation, but it's not perfect at all. This method is already deprecated and will be replaced by clean API calls in my work context. In the next few months we want to provide a full REST API to handle all the load tasks and then completely remove those custom rake tasks.
Via the REST API
This is the preferred way you should consider for automatically populating the CMDB.
Cartoque aims at exposing a clean REST-fulie API. It's not really well-documented nor complete yet, but we're working on it!
Via a Mongodb client on the database itself
We DO NOT recommend this at all. Such method can break the database or some application constraints, and there's no warranty at all that something working today will work with future releases of Cartoque. Be careful!