Question

I have a simple Perl script which runs as a Linux daemon using an infinite loop. It connects to a database every 10 seconds to execute a process.

while (1)
{
    # THIS LINE WILL KILL THE SCRIPT IF IT FAILS
    my $DB=DBI->connect("dbi:Sybase:server=myserver","user","password");
    . . . do something . . .
    sleep (10);
}

I have two questions:

• If the DB is unavailable, how I can keep the script running?
• Can I add an exception handler to send me an email or log an error?

Answer 1

This attempts to connect at 10 second intervals, not every 10 seconds, as William Pursell noted:

while (1)
{
    # THIS LINE WILL KILL THE SCRIPT IF IT FAILS
    my $DB;
    eval { 
        $DB = DBI->connect("dbi:Sybase:server=myserver","user","password");
    };
    if ( my $ex = $@ ) {
        warn $ex;
        next;
    }
    # do something with $DB
    continue {
        sleep 10;
    }
}

See also Object Oriented Exception Handling in Perl, is it worth it? and How can I cleanly handle error checking in Perl?