Bug#308564: [Pkg-nagios-devel] Bug#308564: nagios-pgsql: README.pgsql is (slightly) wrong

[sean finney]

--fUYQa+Pmc3FrFX/N
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

tags 308564 pending
tags 308568 pending
thanks

hi turbo,

nice to hear from you.

On Wed, May 11, 2005 at 07:29:26AM +0200, Turbo Fredriksson wrote:
>   The documentation say that I should load the database BEFORE it's creat=
ed:

fixed in cvs, will include in next upload.

On Wed, May 11, 2005 at 08:23:36AM +0200, Turbo Fredriksson wrote:
> Problem 1a:
>   Also, the readme doesn't state that the resulting file (after
>   the copy and patching) should be made executable. Obvious after
>   the following message but still:
>=20
>   bash: /usr/lib/nagios/plugins/check_nagios_db: /usr/bin/perl: bad inter=
preter: Permission denied

fixed.

> Problem 1b:
>   The readme file states that /etc/nagios/nagios.cgi should be edited
>   to add the nagios_check_command statement. But it seems that it should
>   be the cgi.cfg file:

yup, fixed.

>   Note that the command THERE is '/etc/nagios/check_nagios_db', NOT
>   the stated '/usr/lib/nagios/plugins/check_nagios_db'. The former locati=
on
>   is a must (init script etc uses this - which is wrong in it self, it
>   should extract the 'nagios_check_command' option if set from the config=
).

yes, it should.  ideally, check_nagios_db should be in /usr/lib/nagios,
but unfortunately it has some fields that must be edited for it to be
of any use, which is why i imagine this current setup was recommended.
what we *should* do is seperate out the config into an
/etc/nagios/check_nagios_db.pm (or whatever the appropriate perl config
extension is), and then include that from the script in /usr/lib.
then, as you suggest, one could grep the setting out of the config
file in the init script, which would keep things backwards compatible
for those who already have followed the current instructions.

if you feel like it, please clone/open a new bug about this (this one
will be changed with the documentation fixes).  if you *really* feel
like it, patches are always welcome too :)

> Problem 1c:
>   The readme states that one should test the new script, but first time
>   that fails:
>=20
>   aurora:/usr/share/doc/nagios-pgsql# /usr/lib/nagios/plugins/check_nagio=
s_db
>   Nagios problem: no process identifier found in the database
>   Database handle destroyed without explicit disconnect.
>=20
>   Also, it's not possible to restart nagios. It segfaults after the
>   configuration changes. I'll have to dig deeper (I've forgot :) on
>   why this is happening.

huh.  haven't seen this behavior before.  if it turns out this is
a reproducible problem, you should open a seperate bug about this too.



	sean

--=20

--fUYQa+Pmc3FrFX/N
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFCgpqcynjLPm522B0RArj4AJsGqTs5nhKO+xlT3oqW5feRlcJeMwCfQzYO
7bkMKCmYcfMJZNeaXmtFsEo=
=BimO
-----END PGP SIGNATURE-----

--fUYQa+Pmc3FrFX/N--