Finding discrepancies between data sources
==========================================

There are two data sources here:

	names.passwd	Unix passwd-file format
	names.ldif	LDIF format

They are supposed to contain matching data, but it is suspected that
the two systems have drifted out of sync.

1)	Load the LDIF data into an LDAP server.
	You will either have to create a new suffix dc=example,dc=org
	in the server, or use IDI to map all the entries into an
	existing suffix. If you decide to create the new suffix, you will
	need to load the create-org.ldif file to create the organisation
	entry before trying to load the names.ldif file

2)	Use IDI to generate a list of differences.
	You can assume that the username attribute is a good join field.
	Differences can be expected in any name fields:
		cn
		sn
		displayName
	There may be entries that appear in only one of the data sources.

3)	Consider how you would use IDI to bring the LDAP data into line
	with the passwd-file data.
	This is harder than it first seems because CN is used as part
	of the DN.

Andrew Findlay
27 January 2005

andrew.findlay@skills-1st.co.uk

