Hello, I am using pandora vmware image to monitor several cisco routers through SNMP. Here the time zone is GVT+10 (Brisbane) and I have all configs on vmware image for that:


# date
Wed Nov 19 09:11:52 EST 2008

# hwclock
Wed 19 Nov 2008 09:10:46 AM EST  -0.387166 seconds

# ls -la /etc/localtime
lrwxrwxrwx 1 root root 38 2008-11-04 15:23 /etc/localtime -> /usr/share/zoneinfo/Australia/Brisbane

# cat /etc/php5/apache2/php.ini | grep timezone
date.timezone = Australia/Brisbane

I don't know why but something is going wrong with this function (or probably I am doing something wrong) but monitor just get one data and stop. I was investigating what is happen and the problem is most likely to be here:


# cat  /usr/local/bin/pandora_network | grep UNIX_TIMESTAMP
                              ((tagente_estado.last_execution_try + tagente_estado.current_interval) < UNIX_TIMESTAMP())
                                ((tagente_estado.last_execution_try + tagente_estado.current_interval) < UNIX_TIMESTAMP() )	

This sentence never goes true because “ tagente_estado.last_execution_try” is 54000 seconds ahead than it should be. Then I added this constant (54000) and solved apparently.

# cat /usr/local/share/perl/5.8.8/PandoraFMS/DB.pm | grep "UnixDate" | grep utime
    my $utimestamp = &UnixDate($timestamp, "%s") - 54000; # convert from human to integer  

The problem now is when the host goes down and comes up again all snmp_inc stop to work. If I force the flag of the module (flag=1) it starts again normally. Do you have any idea what I am doing wrong? Kind Regards, Flavio