[check_postgres] bug in txn_wraparound perfdata output

[Cédric Villemain]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Greg Sabino Mullane a écrit :
>> The following lines must be reversed :
>>
>>  $db->{perf} .= " $dbname=$dbtxns";
>>  next SLURP if skip_item($dbname);
> 
> Are you sure this is what you want? By default, we include everything in the
> perf stats, even those that are skipped (e.g. see check_backends). We should
> probably standarize one way or the other. Anyone have strong feelings about
> this? I'm fine with either way.
> 

Well, for my scenarii it is better to have only relevant data in the perfdata.
Using include and exclude rules let me arrange the perfdata and assume it
getting the same elements and order. This way pnp4nagios just push the data in
an rrd.

Looking at check_bloat for example, it is not the same : data is skipped or
analyzed and push in the perfdata if needed.


So we have at least 2 ways :

  1/ include everything everytime in the perfdata *even if* the content is not
analyzed (so don't throw alert).

  2/ only include data which can lead to an alert.

perhaps 3 :

  3/ add an output pnp4nagios (so free to get specfic rules).


I prefer 2 (or 3 but this one is less elegant).








- --
Cédric Villemain
Administrateur de Base de Données
Cel: +33 (0)6 74 15 56 53
http://dalibo.com - http://dalibo.org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAknRyeIACgkQo/dppWjpEvxINwCfReEdFIaSV81x9cNvpL7VBxkC
A3UAoI0aPdDpbS6IEyJ3wGyRgYhdrTvN
=2siJ
-----END PGP SIGNATURE-----