It appears that MySql is doing funny things to SIGALRM.

No, it doesn't. It's the first bullet. The alarm signal is received. C calls are not interruptible, so calling the handler is delayed until the C code returns. alarm is working as expected.

1) Is this problem fixable?

You can interrupt C code by using unsafe signals, but it can leave the process in a weird state.

2) The whole purpose of this exercise is to find a way to get out of the MySql operation if it takes longer than 5 sec. to do MySql.

Perhaps you should start a thread with that question.

You can interrupt C code by using unsafe signals,

So Re: Perl Alarm Not Working isn't such a dumb suggestion after all.

While subjecting oneself to heisenbugs and seg faults willingly is not necessarily dumb, it's definitely not wise.

       C calls are not interruptible

Hmm...?? How do we explain the followings:

I made a demostration delibrately make MySql execute to a Maximum time of 5 sec, then set a timeout alarm of 3 sec in eval{} to prevent MySql execution more than 3 sec.

To my surprise, timeout alarm in eval{} stops MySql execution exactly at 3 sec. COSISTENTLY

The followings is the demonstration script:


#!/usr/bin/perl 

$max_time = 5;
$timeout = 3;
use DBI;

print "Content-type: text/html\n\n";
sub do_counter {

    $time1 = time;
    $time2 = time;
    $counter = 0;
    $time_ran = $time2 - $time1;

        $dbh=DBI->connect("dbi:mysql:$database:localhost","username","
+password");
    $query="SELECT name FROM test";
    $sth=$dbh->prepare($query);     
        $sth->execute();     
        while ( @row = $sth->fetchrow_array ) {

        while($time_ran <= $max_time) {

            $conter++;
            $time2 = time;
            $time_ran = $time2 - $time1;
        } 
                                        
    }

    $sth->finish;

    $dbh->disconnect ||die("Couldn't disconnect to database!\n");

}


eval {        
        local $SIG{ALRM} = sub { die "alarm\n" }; 
        alarm $timeout;        
        &do_counter;        
        alarm 0;    
}; 

if($@) {        
        print "$@, Time Ran: $time_ran, Counter: $conter\n";
    exit;
      
}

print "Time Ran: $time_ran, Counter: $conter\n";
exit;
[download]


As you can see, I delibrately make a counter loop inside MySql's while ( @row = $sth->fetchrow_array ) loop. The counter loop can run to a $max_time of 5 sec. before it stops.

In the eva{} I set a timeout alarm of 3 sec. to stop MySql execution from going over 3 sec.

When I ran the above script, I got the following results consistently:

alarm, Time Ran: 3, Counter: 3341554
alarm, Time Ran: 3, Counter: 3416456

These results show that alarm in eval{} has stopped MySql execution in 3 sec.

Can we not say then that alarm has interrupted MySql C Calls (e.g. calls to the db driver), contrary to what you said?

You can do the same demostration with the above script by doing the followings:

1) Create a table TEST with a single column: name CHAR(30) NOT NULL
2) Enter 3 names, like John, Paul, Mary in the table.
3) Run the above script.

Can we not say then that alarm has interrupted MySql C Calls (e.g. calls to the db driver), contrary to what you said?

No.

        while ( @row = $sth->fetchrow_array ) {

        while($time_ran <= $max_time) {

            $conter++;
            $time2 = time;
            $time_ran = $time2 - $time1;
        } 
                                        
    }
[download]

isn't C.

Can we not say then that alarm has interrupted MySql C Calls (e.g. calls to the db driver), contrary to what you said?

No, pay attention , unsafe signals will interrupt c calls. Refer to documentation, its what we do.