check_ldaps fails to run successfully inside nagios schedule

Nagios Nagios Plugins support
1

I am running nagios 1.2 on a 2.4.21-27.ELsmp i686 host without trouble.
We are needing to migrate off that host and to a host running
2.6.9-34.ELsmp x86_64 host. I’d like to upgrade nagios as well, but time won’t allow
currently, so I’m doing a straight port of 1.2 onto the new host for the
time being.

My issue is that the plugin check_ldaps works on the old host, but not
the new one. It gets even more strange. I can get the command line,
any user, to run check_ldaps successfully, but the nagios binary
continues to launch failed check_ldaps connections.

I’ve updated /etc/openldap/ldap.conf to include TLS_CACERT line to
indicate location of generic cert to use. I however don’t have much
knowledge about the ldap innards making this extra difficult to
troubleshoot.

What I do know is that any user on new host can run check_ldaps and it
works, but it shows up as failed in the new nagios browser. The nagios
binary has to be started as root

Ideas?

Dave

2

Found the fix for those wondering. For whatever reason, the check_ldaps would not bind using the IP address whilst with the hostname it worked beautifully( using $HOSTALIAS$ instead of $HOSTADDRESS$ <–this being the diff from nagios binary to command line…i was running cmd line with hostname

Also note that using double quotes around the ldap query parms proved a problem when the DN had a space in the name; using single ticks around that fixed that issue as well:

   command_line    $USER1$/check_ldaps -H $HOSTALIAS$ -b o=company.com -p 636 -D 'cn=mytest,ou=IT Tester, o=company.com' -P $USER5$ -w 150 -c 300 -t 300