[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: SQL backend specific options - description?

Hello Heiko!

Heiko Nardmann wrote:

> Is there any document describing the SQL backend specific options
> already?

install and concept files contain partial information about several of
them... most of them were considered self-explanatory, and
you can find illustrations on most of directives in examples for different

well, let's try to post some more explanations here (blame my English again
if they are not understandable ;):

- dbname - the name of ODBC datasource to use
- dbhost, dbuser, dbpasswd - are generally unneeded, because this
information is already taken from datasource.  Use them if you need
override datasource settings. Also, several RDBMS' drivers tend to require
explicit passing of user/password, even if those are given in datasource.
-  subtree_cond - a SQL where-clause template used to form subtree search
condition. it may differ from one SQL dialect to another (see samples)
-  oc_query, at_query, insentry_query, delentry_query - SQL query templates
for loading schema mapping metainformation, adding and deleting entries to
ldap_entries, etc. All these and subtree_cond have default values, which
can be seen in sources, or in log output when slapd starts with "-d 5" or
- upper_func - if given, is interpreted as name of function that converts
given value to uppercase. this is used for CIS matching when RDBMS is case
sensitive. I.e., for Oracle this is set to "UPPER" (see sample)

to the list:

there are some more to come, inspired by Postgres port - with them back-sql
will become entirely customizable to SQL dialects in all it's parts
(probably there IS some _weird_ dialect that won't fit principally, but
that is not back-sql's fault - use some compatibility level software such
as EasySoft provides, to normalize SQL between your RDBMSes, or write new
standard ;)

BTW, there are some more long-requested features which would be of general
interest (like IP address available for substitution in queries, and
password setting exop support), which are fairly easy to implement, but for
which I have no time currently. If someone can help with patches - it would
be very nice.

WBW, Dmitry