Re: back-config design considerarions - Admin Guide fodder

[Michael StrÃder <michael@stroeder.com>]

Quanah Gibson-Mount wrote:
> 
> --On Thursday, July 28, 2005 7:18 PM +0200 Michael StrÃder
> <michael@stroeder.com> wrote:
> 
>> Michael StrÃder wrote:
>>
>>> If I'm using option -f slapd.conf and -F configdir/ together which
>>> config data is authorative?
>>> [..]
>>> My conclusion:
>>> Drop -f slapd.conf completely in 2.3.x and rather develop good setup
>>> tools...
>>
>> Well, same question for file DB_CONFIG and attribute 'olcDbConfig'...
> 
> I can't see -f going away until at least *everything* supports the new
> config backend, which it doesn't.
> 
> Besides which point, when I do testing of various setups, I use -f
> extensively to switch between configuration files on the same host,
> which is a lot simpler than mucking around with various config
> backends.  Other people may want to keep their slapd.conf files for
> things like version control, etc, as well.

Quanah, I do see all the advantages of slapd.conf mentioned above. But I
also see the issues with inconsistencies and user confusion. Arent' you
then questioning usefulness of back-config in general?

Also the admin guide has to describe both methods not making it more
readable for beginners (although Howard did a great job updating it).

BTW: Testing various setups by using -F with different directories is as
easy as using -f. And you can check in the configdir/ tree to CVS as
well without problem.

Furthermore if in the future ACLs and access/audit logging is
implemented for back-config you have a much more powerful instrument to
track configuration changes than CVS for slapd.conf.

=> One has to decide which route to go and after that one has to follow
that route consequently to avoid inconsistencies and double effort.

Ciao, Michael.