On Sun, 17 Jun 2012, Bjoern Wuest wrote: ...
However, setting up the mail system (dovecot + postfix) I encountered a problem new to me. When I try to bind as a "normal" user (here: bjoern) to LDAP it fails with wrong credentials. I can assure that I did not mistyped the password (tried multiple times). Login to the Linux system and samba with same credentials (i.e. bjoern and his password) works.
Here is the part of syslog I expect to be the cause:
Jun 17 19:36:45 server slapd[23241]: <<< dnPrettyNormal: <cn=bjoern,ou=Users,dc=domain,dc=my>, <cn=bjoern,ou=users,dc=domain,dc=my> Jun 17 19:36:45 server slapd[23241]: conn=1003 op=0 BIND dn="cn=bjoern,ou=Users,dc=domain,dc=my" method=128 Jun 17 19:36:45 server slapd[23241]: do_bind: version=3 dn="cn=bjoern,ou=Users,dc=domain,dc=my" method=128 Jun 17 19:36:45 server slapd[23241]: Jun 17 19:36:45 server slapd[23241]: ==> hdb_bind: dn: cn=bjoern,ou=Users,dc=domain,dc=my Jun 17 19:36:45 server slapd[23241]: bdb_dn2entry("cn=bjoern,ou=users,dc=domain,dc=my") Jun 17 19:36:45 server slapd[23241]: daemon: epoll: listen=8 active_threads=0 tvp=zero Jun 17 19:36:45 server slapd[23241]: => hdb_dn2id("cn=bjoern,ou=users,dc=domain,dc=my") Jun 17 19:36:45 server slapd[23241]: daemon: epoll: listen=9 active_threads=0 tvp=zero Jun 17 19:36:45 server slapd[23241]: daemon: epoll: listen=10 active_threads=0 tvp=zero Jun 17 19:36:45 server slapd[23241]: daemon: epoll: listen=11 active_threads=0 tvp=zero Jun 17 19:36:45 server slapd[23241]: <= hdb_dn2id: get failed: DB_NOTFOUND: No matching key/data pair found (-30987)
In my experience, that sort of error from the DB library usually means a change to the indexing or schema was made without reindexing and/or dumping and reloading.
If you're confident that's not the case here (how confident?), then have you compared that log output to the log output of a successful login?
Philip Guenther