Apache::AuthzDBI(3pm)

NAME

Apache::AuthzDBI - Authorization via Perl's DBI

SYNOPSIS

 # Configuration in httpd.conf or srm.conf:

 PerlModule Apache::AuthzDBI

 # Authorization in .htaccess:

 AuthName DBI
 AuthType Basic

 #authorize via DBI
 PerlAuthzHandler Apache::AuthzDBI

 PerlSetVar Auth_DBI_data_source   dbi:driver:dsn
 PerlSetVar Auth_DBI_username      db_username
 PerlSetVar Auth_DBI_password      db_password
 #DBI->connect($data_source, $username, $password)

 PerlSetVar Auth_DBI_pwd_table     users
 PerlSetVar Auth_DBI_uid_field     username
 PerlSetVar Auth_DBI_grp_field     groupname
 #SELECT grp_field FROM pwd_table WHERE uid_field=$user

 require user   user_1  user_2 ...
 require group group_1 group_2 ...

The AuthType is limited to Basic. You may use one or more valid require lines. For a single require line with the tokens valid-user or with distinct user names it is sufficient to use only the AuthenDBI module.

DESCRIPTION

This module allows authorization against a database using Perl's DBI. For supported DBI drivers see:

 http://www.hermetica.com/technologia/DBI/

When the authorization handler is called, the authentication has already been done. This means, that the given username/password has been validated.

The handler analyzes and processes the requirements line by line. The request is accepted only if all requirement lines are accepted.

In case of one or more user-names, they are compared with the given user-name until the first match. If there is no match and the authoritative directive is set to 'on' the request is rejected.

In case of one or more group-names, all groups of the given user-name are looked up in the cache. If the user is not found in the cache, the groups are requested from the database. A comma separated list of all these groups is put into the environment variable REMOTE_GROUPS. Then these groups are compared with the required groups until the first match. If there is no match and the authoritative directive is set to 'on' the request is rejected.

In case of 'valid-user' the request is accepted.

In case the authorization succeeds, the environment variable REMOTE_GROUP is set to the group name, so scripts that are protected by AuthzDBI don't need to bang on the database server again to get the group name.

The SQL-select used for retrieving the groups is as follows (depending upon the existence of a grp_table):

 SELECT grp_field FROM pwd_table WHERE uid_field = user
 SELECT grp_field FROM grp_table WHERE uid_field = user

This way you can have the group-information either in the main users table, or you can use an extra table, if you have an m:n relationship between users and groups. From all selected groups a comma-separated list is build, which is compared with the required groups. If you don't like normalized group records you can put such a comma-separated list of groups (no spaces) into the grp_field instead of single groups.

If a grp_whereclause exists, it is appended to the SQL-select.

At the end a CleanupHandler is initialized, which skips through the groups cache and deletes all outdated entries. This is done after sending the response, hence without slowing down response time to the client. The default cache_time is set to 0, which disables the cache, because any user will be deleted immediately from the cache.

LIST OF TOKENS

• Auth_DBI_data_source: The data_source value should begin with 'dbi:driver_name:'. This value (with the 'dbi:...:' prefix removed) is passed to the database driver for processing during connect.

• Auth_DBI_username: The username argument is passed to the database driver for processing during connect.

• Auth_DBI_password: The password argument is passed to the database driver for processing during connect.

• Auth_DBI_grp_table: Contains at least the fields with the username and the groupname.

• Auth_DBI_uid_field: Field-name containing the username in the Auth_DBI_grp_table.

• Auth_DBI_grp_field: Field-name containing the groupname in the Auth_DBI_grp_table.

• Auth_DBI_grp_whereclause: Use this option for specifying more constraints to the SQL-select.

• Auth_DBI_authoritative < on / off>: Default is 'on'. When set 'on', there is no fall-through to other authorization methods if the authorization check fails. When this directive is set to 'off', control is passed on to any other authorization modules. Be sure you know what you are doing when you decide to switch it off.

• Auth_DBI_casesensitive < on / off >: Default is 'on'. When set 'off', the entered userid and password is converted to lower case.

• Auth_DBI_cache_time Default is 0 = off. When set to any value > 0, the groups of all users will be cached. After finishing the request, a special handler skips through the cache and deletes all outdated entries (entries, which are older than the cache_time).

CONFIGURATION

The module should be loaded upon startup of the Apache daemon. It needs the AuthenDBI module for the authentication part. Note that this needs mod_perl-1.08 or higher, apache_1.3.0 or higher and that mod_perl needs to be configured with

  PERL_AUTHEN=1 PERL_AUTHZ=1 PERL_CLEANUP=1 PERL_STACKED_HANDLERS=1.

Add the following lines to your httpd.conf or srm.conf:

 PerlModule Apache::AuthenDBI
 PerlModule Apache::AuthzDBI

PREREQUISITES

For Apache::AuthzDBI you need to enable the appropriate call-back hooks when making mod_perl:

  perl Makefile.PL DO_HTTPD=1 PERL_AUTHEN=1 PERL_AUTHZ=1 PERL_CLEANUP=1 PERL_STACKED_HANDLERS=1.

AUTHORS

• mod_perl by Doug MacEachern <[email protected]>

• DBI by Tim Bunce <[email protected]>

• Apache::AuthzDBI by Edmund Mergl <[email protected]>

COPYRIGHT

The Apache::AuthzDBI module is free software; you can redistribute it and/or modify it under the same terms as Perl itself.