Author: arkurth
Date: Wed Mar 18 17:45:02 2009
New Revision: 755673

URL: http://svn.apache.org/viewvc?rev=755673&view=rev
Log:
VCL-110
Removed 2 second delay in vcld's loop. It may have caused very rare timing 
problems with imaging.

Commented out 2 notify() messages in the REAPER() sub. They were only needed 
for troubleshooting exit status problems.

Modified:
    incubator/vcl/trunk/managementnode/bin/vcld

Modified: incubator/vcl/trunk/managementnode/bin/vcld
URL: 
http://svn.apache.org/viewvc/incubator/vcl/trunk/managementnode/bin/vcld?rev=755673&r1=755672&r2=755673&view=diff
==============================================================================
--- incubator/vcl/trunk/managementnode/bin/vcld (original)
+++ incubator/vcl/trunk/managementnode/bin/vcld Wed Mar 18 17:45:02 2009
@@ -318,12 +318,6 @@
                                # Add the reservation ID to be processed to the 
hash
                                $request_info{RESERVATIONID} = $reservation_id;
 
-                               # Wait for a short amount of time before making 
new child
-                               # This is to give processes a chance to start 
before subsequent processes
-                               #   check for conflicts such as overlapping 
computer reservations
-                               notify($ERRORS{'DEBUG'}, $LOGFILE, "sleeping 
for 2 seconds before updating state to pending");
-                               sleep 2;
-
                                # Update the request state to pending, 
laststate to next state
                                # Pending is set now so vcld doesn't try to 
process it again
                                # The previous state is already in the hash as 
the laststate value
@@ -748,7 +742,7 @@
                $dead_pid = $wait_pid;
                
                # Assemble a string containing the dead process info
-               notify($ERRORS{'DEBUG'}, 0, "process reaped: pid: $dead_pid, 
\$?: $status_save, exit status: $child_exit_status");
+               #notify($ERRORS{'DEBUG'}, 0, "process reaped: pid: $dead_pid, 
\$?: $status_save, exit status: $child_exit_status");
                
                # Check if the child PID hash contains the pid of the process 
which just died
                if (exists $child_pids{$dead_pid}) {
@@ -767,7 +761,7 @@
        # Set the special $? variable back to the exit status of the child 
which died
        # This is useful when utilities such as SSH are run in other places in 
the code
        # The code which called the utility can check the exit status to see if 
it was successful
-       notify($ERRORS{'DEBUG'}, 0, "setting \$? to $status_save, exit status: 
$child_exit_status");
+       #notify($ERRORS{'DEBUG'}, 0, "setting \$? to $status_save, exit status: 
$child_exit_status");
        $? = $status_save;
        
        return;


Reply via email to