joeymac:

The only things I can think of are:

• There may be an alarm set elsewhere in the code.
• There may be resource limits on the user account that's shutting the job down.
• There may be a connection timeout (firewall or some such) that's closing the connection and shutting the job down.

...roboticus

When your only tool is a hammer, all problems look like your thumb.

Thanks for the quick response! There are no alarms coded in the script. Your second suggestion is something that I am not sure about. I will look into this. The third suggestion I believe I accounted for with a second nested infinite loop that tries to reconnect (using Net::FTP) until connection is reestablished. It then exits this loop with a last statement and returns to the overall infinite loop.

 
for(;;) {...

my $host = something;
my $ftp = Net::FTP->new($host, Timeout => 30);
  if (defined($ftp)) {
      my $TODAY = time;
      print LOG "Established ftp connection with $host\n";
      print LOG "at: ", scalar(localtime($TODAY)), "\n";
   } else {
      for (;;) {
       print LOG "Unable to make connection with $host\n";
       print LOG  "Try to reconnect in 5 secs to $host\n";
       sleep 5;
       my $ftp = Net::FTP->new($host, Timeout => 60);
       if (defined($ftp)) {
         my $TODAY = time;
         print LOG  "Established ftp with $host\n";
         print LOG  "at: ", scalar(localtime($TODAY)), "\n";
         last;
       }
   }
 }
...}
[download]

What is the last message that's printed to LOG?

That might help in finding the call that's causing the script to exit.

   for (;;;)
[download]

  while(1)
[download]

Options?

You could set up the script to run under some process monitor or other. There are quite a few around. For a bit of fun you could write your own (but who monitors the monitor?).

If you wrap the whole thing in an eval you maybe able to trap the condition that causes the exit and resume the loop.

What I did was write a quick script to send STDERR to a log file. Things went astray again and this is the message that I got.

 fileparse(): need a valid pathname at ftpData.pl line 312

I am using fileparse (File::Basename) to disregard any file extensions in directory comparisons and to remove the existing extension when the file is copied and renamed with a different extension, like so:

 #Disregard file extension in comparison:
foreach my $item (@wrkDirList) {
my ( $fileName, $filePath, $fileExt ) = fileparse($item, qr/\.[^.]*/);
    $item = $fileName;
}
[download]

and

 foreach $item (@filesToRetrv) {
    my ( $fileName, $filePath, $fileExt ) = fileparse($item, qr/\.[^.]
+*/);
        if (rename($item, $fileName)) {
           push (@filesToMove, $fileName);
        } else {
           print LOG "Rename failed for $localName to $fileName\n";
        }

    }
[download]

So, and correct me if I am wrong, somewhere along the line it looks like fileparse is getting passed an empty filehandle? I'm not really sure how this could happen. The STDERR message seems to be occurring at the second instance of fileparse (the one working on @filesToRetrv). Here is a little snippet with info about this array and what goes into it:

foreach $fileToGet (@matches) {
  my $remoteFileSize = $ftp->size($fileToGet);
  $localFileName = "$fileToGet".".xxx";
  $ftpReturnVar =  $ftp->get($fileToGet, $localFileName);
  $localFileSize = (stat "$workingDir/$localFileName")[7];
    if ($remoteFileSize == $localFileSize) {
        push (@filesToRetrv, $ftpReturnVar);
    } else { ... }
}
[download]

Any ideas? Thanks for the help!

One thing lest the PBP fundamentalists get you - always use a lexical variable in foreach loops:

    foreach my $fileToGet (@matches) {
    ...
[download]

What happens if $ftp->get fails? You may want to check whether you get a valid return value e.g.

  $ftpReturnVar =  $ftp->get($fileToGet, $localFileName);
  defined $ftpReturnVar or next;
[download]

for ( ... ; ... ; ... ) is not a counting loop. It never creates a counter. It's really a while loop in disguise (in terms of function).

What's the process's exit code? It will tell you which signal killed the process if it was a signal (e.g. SIGSEGV, SIGCHLD, SIGPIPE), and the exit code will likely be an errno if the process didn't die from a signal. (perl -E'say $!=123;' to find out what errno 123 means.)

while(1){... do stuff...}
[download]

while (1) and for (;;) (which is also quite traditional) result in the same code.

$ perl -MO=Concise,-exec -e'for (;;) { f() }'
1  <0> enter 
2  <;> nextstate(main 3 -e:1) v:{
3  <{> enterloop(next->8 last->9 redo->4) v
4  <;> nextstate(main 1 -e:1) v
5  <0> pushmark s
6  <#> gv[*f] s/EARLYCV
7  <1> entersub[t3] vKS/TARG,1
8  <0> unstack v
-e syntax OK

$ perl -MO=Concise,-exec -e'while (1) { f() }'
1  <0> enter
2  <;> nextstate(main 3 -e:1) v:{
3  <{> enterloop(next->8 last->9 redo->4) v
4  <;> nextstate(main 1 -e:1) v
5  <0> pushmark s
6  <#> gv[*f] s/EARLYCV
7  <1> entersub[t3] vKS/TARG,1
8  <0> unstack v
-e syntax OK

$ diff -u \
      <( perl -MO=Concise,-exec -e'while (1) { f() }' 2>&1 ) \
      <( perl -MO=Concise,-exec -e'for (;;)  { f() }' 2>&1 ) \
   && echo same
same
[download]

No particular reason why I used

 for(;;) {...}

instead of

 while(1) {...}

It is my understanding that the two are pretty much interchangable.