LDAP database corrupyion (ITS#615)

[tsingh@dominomail.com]

We are running Openldap 1.2.7.2 for authentication of users accessing our dialin
users and for the mail service.
we have a primary and secondary ldapserver with replication running between
them.
yesterday there was a problem with the slurpd.therefore to isolate the problem,
we ran the slurpd and slapd daemon in debugging mode with the following syntax:

/usr/sbin/slapd -f /etc/openldap/slapd.conf -d 68 ( on primary and secondary
servers)
/usr/sbin/slurpd -f /etc/openldap/slapd.conf -d 68 (on primary server)
the ldapsearch on both the servers yielded the result as desired.on exiting
the debugging mode for the slurpd by pressing:
cntl+c,
the slurpd replication process was stopped.similarly the slapd daemon was stopped.

we then tried to run the ldap daemon by running:
/etc/rc.d/init.d/ldap start
the slapd and the slurp daemons started and were confirmed by :
ps -ef | grep slapd
ps -ef | grep slurpd
however now the ldapsearch gave no result nor any errors.
i suspected ldap database corruption.
on checking the ldap database directory, there were only 3 files and no ldi2entry.dbb,objectclass.dbb
etc.
also the secondary server also lacked the files as mentioned.
we had to manually create the database.
however we are unable to find any plausible explanation for the above incident.

since we are running an isp with a user database of over 30,000 users, we cannot
afford to have this failure in future.
i d require any help on the above incident ASAP.
also note that the access to the servers is highly restricted and no one could
havedeleted the database.
thanks in advance to all those who can help 
and regards
tajeshwar singh
  
Free Web Mail for Notes Professionals http://www.dominomail.com