Service URL Monitoring

[MOHAMMAD1412]

Hi All,

We are monitoring Service URL status using Nagios. Currently application has been upgraded, once after the upgrade I can see lot of alerts getting triggered stating that "WEB SITE is having problems".

Below is the command used to check the URL Status

/usr/bin/curl -A "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 6.0)" --basic "http://xx-xxxx-xxx.xxx.xxx.com" | /bin/grep xxxx

Upgraded application version supports IE 11, so I have corrected the above command with the new IE version but still its showing in Warning status.

/usr/bin/curl -A "Mozilla/4.0 (compatible; MSIE 11.0; Windows NT 6.0)" --basic "http://xx-xxxx-xxx.xxx.xxx.com" | /bin/grep xxxx

Please provide your comments.

Thanks In Advance

[rkennedy]

It doesn't look like you're using our check_http plugin. What are you using for warning / critical thresholds?

[MOHAMMAD1412]

Yes check_http plugin is not used, we are using Perl script for check the Application URL status.

[rkennedy]

As you're using a custom script, we do not know everything that is going into it. Can you post the perl script here?

[MOHAMMAD1412]

Hi,

Please find the required parameters

Service.cfg
define service{
use generic-service
host_name xx-xxxxx-xxx
service_description HTTP-TEST
normal_check_interval 30
contact_groups xxxx-xxxx
notification_interval 60
notification_options c,r,w,u
check_command generic-http-check!test
}

define command{
command_name generic-http-check
command_line /usr/local/nagios/libexec/generic-http-chk.pl -H $HOSTNAME$ -f $ARG1$
}

Command used by the script - /usr/bin/curl -A "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 6.0)" --basic "http://xx-xxxxx-xxx.xxx.xxx.xxx" | /bin/grep Test

Script

#! /usr/bin/perl -w
#
# usage:
#
use POSIX;
use strict;
use Getopt::Long; declaration of Getopt command
use vars qw($opt_H $opt_f );
use vars qw($PROGNAME);

use lib "/usr/local/nagios/libexec" ;
use utils qw($TIMEOUT %ERRORS &print_revision &support &usage);

sub print_help ();
sub print_usage ();

$PROGNAME = "generic-http-chk.pl";

$ENV{'PATH'}='';
$ENV{'BASH_ENV'}='';
$ENV{'ENV'}='';

Getopt::Long::Configure('bundling');
GetOptions
("H=s" => \$opt_H, "hostname=s" => \$opt_H,
"f=s" => \$opt_f, "filter=s" => \$opt_f);


($opt_f) || ($opt_f = shift) || ($opt_f = "FILTER OUT ALL BUT");


use strict;
my($status,$cmd,$output,$junk);
my @output;


$cmd = "/usr/local/bin/curl.$opt_f";

chop(@output=`$cmd`);
$status = $?;
$status = $?/256; # To get actual exit value, divide by 256
($output)=@output;
#print "$status\n";

if ($status==0) {
print "$opt_f WEB SITE is OK\n";
exit 0;
}
elsif ($status==1) {
print "$opt_f WEB SITE is having problems\n";
exit 1;
}
elsif ($status==2) {
print "$opt_f WEB SITE is in big trouble\n";
exit 2;
}
else {
print "@output\n";
exit 3;
}

Thanks

[rkennedy]

I tried getting the script you posted to work, and was unable to. As you're using a custom plugin, we can't support everything.

With that said, the default check_http plugin does support a useragent with -A, will this plugin not work for you?

-A, --useragent=STRING
String to be sent in http header as "User Agent"

[MOHAMMAD1412]

Hi All,

Updated the script resolved the issue.

Please close the thread.

Thanks

[rkennedy]

Great to hear, marking this as resolved and closing it out.