On Thu, Oct 15, 2009 at 09:46:28PM +0200, Jan Wagner wrote: > > Now, why didn't check_ldap communicate that? Because it has this > > in the code (plugins/check_ldap.c): > > > > /* bind to the ldap server */ > > if (ldap_bind_s (ld, ld_binddn, ld_passwd, LDAP_AUTH_SIMPLE) != > > LDAP_SUCCESS) { > > /*ldap_perror(ld, "ldap_bind"); */ > > printf (_("Could not bind to the ldap-server\n")); > > return STATE_CRITICAL; > > } > > > > How hard was it to put that ldap_perror() string in the printf'ed > > error message? :( > > > > A quick grep for ldap_perror shows that there are other occurences of the > > same problem in the same file. > > looking into the actual code and into the git tree, indicates, that there > should be changes about this issue. With 1.4.10 should be introduced[1] a > verbose option, which looks like what you want. Can you please have a look > into it and give feedback.
Thanks for the hint. On the face of it, it looks better. But, it actually seems to be have been done a bit sloppy once again - the plug-in development guidelines http://nagiosplug.sourceforge.net/developer-guidelines.html#AEN39 say that multi-line output needs to be made only when the verbose option is set to >= 2. I believe also that the BCP is to add extra information in the lines *after* the first single-line summary, not before it. As for the Debian bug, it's also relevant that the -v option is not used in any of the command definitions in the default /etc/nagios-plugins/config/ldap.cfg that is shipped with the package (at least in the lenny version that I'm looking at now). FWIW I still think that actual LDAP error info belongs to the summary line, or at least that the package config should default to what is defined as -v 1 or better. People who don't want any extra info can always skip long output by using only $HOSTOUTPUT$/$SERVICEOUTPUT$ in the notification command, and skip $LONGHOSTOUTPUT$/$LONGSERVICEOUTPUT$. -- 2. That which causes joy or happiness. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org