Starting and Stopping Directory Instances

The following scripts should be used to start and stop directory instances:

~iplanet/local/ldap/scripts/directory_start.ksh ~iplanet/local/ldap/scripts/directory_stop.ksh

Usage:

~iplanet/local/ldap/scripts/directory_start.ksh <Inst> e.g. ~iplanet/local/ldap/scripts/directory_start.ksh Test

It is important to utilize the stop|start scripts as they establish runtime environment necessary to the correct functioning of the kerberos plugin installed in the various directory instances.

Power Broker can also be used to stop and start directory instances though it is no longer necessary:

pbrun Teststart-slapd pbrun Teststop-slapd

When Power Broker is invoked, it executes the directory_start|stop.ksh scripts referenced above. Power Broker support exists is because we used to utilize ports < 1024 which require root access to start. That is no longer necessary; all directory instances are now associated with ports above 1024.

System startup scripts also invoke the directory_start|stop.ksh scripts.

Before restarting a stopped directory, attempt to determine the reason the directory is down. The best place to look for possible error conditions are the iPlanet errors* logs in the following locations:

/e01/iplanet/servers/slapd-<server><Inst>/logs/

Depending on what caused the directory to shut down, startup can take a substantial amount of time while the directory instance is being "recovered".

After starting a directory, it pays to confirm that the director has started successfully and is serving requests. The directory_start.ksh script performs as 'ps -ef' with grep for the specific directory process. Directory processs are named as follows:

slapd-<server><Inst>

In addition, you can perform a quick directory search using the following command:

ldap_query<Inst> "(ou=routing)" dn

Replace <Inst> with Test, Prep or Prod (case sensitive).

#############################################
20080429 sxelm

(eof)