Amanda-Catalog

Using the database:

  - See the amanda-catalog man page
    (http://wiki.zmanda.com/man/beta/amanda-catalog.7.html)
  - Modify the configuration file of ALL backup set:
    Add the following to amanda.conf of ALL backup set (for a sqlite catalog):

	# define a catalog
	define catalog "SQLite" {
	  comment "sqlite catalog for amanda"
	  plugin "SQLite"
	  property "dbname" "/var/lib/amanda/database"
	}
	# tell which catalog to use
	catalog "SQLite"


  - Build the catalog
    For each backup set:

	amcatalog CONF create

This operation read the tapelist and the log.TIMESTAMP files and add the
information in the catalog.


All amanda commands now use the catalog instead of the text files.
The log.TIMESTAMP are still created, they are used by amreport and can
be used to recreate the database if needed.
The tapelist file is still maintained, but amanda never read it, you must
never write to it.

You must use the amadmin/amlabel commands or the REST API to do modification
to the catalog/tapelist.


Getting the list of dump available for recovery

  - The 'amadmin find' command do not use the catalog, it still use the tapelist and log.TIMESTAMP files, so you should not use it.
  - Use the 'amcatalog CONF dump' or 'amcatalog CONF part' to list all dumps available for recovery. They are faster especially if you do a query for a single DLE.
  - Using the REST API (http://wiki.zmanda.com/pod/beta/Amanda/Rest/Dumps.html) is the better choice