Openldap bug on centos6,4

Openldap bug on centos6,4

When you are setting up the new openldap server, specifically with the following combination.

[root@centos6-01 openldap]# cat /etc/centos-release
CentOS release 6.4 (Final)

[root@centos6-01 openldap]# uname -r
2.6.32-358.6.1.el6.x86_64

[root@centos6-01 openldap]# rpm -qa | grep openldap-servers
openldap-servers-2.4.23-32.el6_4.1.x86_64

Most of the problem that you will hit when you are trying to add a first ldap entry from a ldif file was "ldap_bind invalid credentials (49)". Luckily, when you are doing a search from google, you should be able to find this link to help you to solve the problem.  http://www.ezylinux.com/en/red-hat-6-how-to-fixes-ldapadd-ldap_bind-invalid-credentials-49/. However, right after you have applied the commands, I just found "ANOTHER" bug on the daemon script as depicted as below. Even though the daemon script yielded an error message, but the slapd process will still spawned out, which I think it is a bit ugly.

[root@centos6-01 openldap]# /etc/init.d/slapd start

ln: accessing `/var/run/openldap/slapd.pid': No such file or directory

[root@centos6-01 openldap]# pidof slapd

1857

[root@centos6-01 openldap]# ps -eaf | grep slapd

ldap      1857     1  0 15:07 ?        00:00:00 /usr/sbin/slapd -h  ldap:/// ldapi:/// -u ldap

I keep on finding for bug list reported from the redhat bugzilla, but I have no luck. So, I decided to hack the daemon script, /etc/init.d/slapd. In this script, you will find some lines sound as below.

 # Start daemons.

        echo -n $"Starting $prog: "

        daemon --pidfile=$pidfile --check=$prog $ulimit ${slapd} -h "\"$harg\"" -u ${user} $OPTIONS $SLAPD_OPTIONS

        RETVAL=$?

        if [ $RETVAL -eq 0 ]; then

                touch $lockfile

                ln $slapd_pidfile $pidfile

        fi

So, this is my hacked version. Please pay more attention on the highlighted lines.

 # Start daemons.

        echo -n $"Starting $prog: "

        daemon --pidfile=$pidfile --check=$prog $ulimit ${slapd} -h "\"$harg\"" -u ${user} $OPTIONS $SLAPD_OPTIONS

        pidof ${slapd} > $pidfile

        RETVAL=$?

        if [ $RETVAL -eq 0 ]; then

                touch $lockfile

                #ln $slapd_pidfile $pidfile

                ln $pidfile $slapd_pidfile

        fi

With this, I can start/stop my daemon with no problem.

[root@centos6-01 cn=config]# /etc/init.d/slapd start

Starting slapd:                                            [  OK  ]

[root@centos6-01 cn=config]# /etc/init.d/slapd stop

Stopping slapd:                                            [  OK  ]

I am not sure if Centos community aware of this bug. But, I am just sharing it out, in hoping that the bug will be fixed in future.