Hello,

I have written a perl script which archives (tar + zip) various directories on Windows 2008 R2 64bit. The server has 8GB of memory.

This is how I get all files which need to be archived:

# Get all export and log files and command file to tar it.
local *push_files = sub { push(@all_files,$File::Find::name); LOG_MESS
+AGE ("### ARCHIVING $File::Find::name\n"); }; 
# nested sub needed as use warnings complains if not
# this from module file::find
find(\&push_files, ($par_COMMAND_TXT,$par_EXPORTTOOL_OUTPATH,$par_EXPO
+RTTOOL_LOG));
[download]

This is how I do the archiving:

my $tar = Archive::Tar->new();
$tar->add_files(@all_files) or die LOG_MESSAGE("!!! Can't run program:
+ $!\n");
# This is needed for older Perl version => otherwise "COMPRESS_GZIP" B
+areword warning
no strict;
$tar->write( "$ARCHIVE_FILE", COMPRESS_GZIP) or die LOG_MESSAGE("Can't
+ run program: $!\n");
LOG_MESSAGE ("### Archiving finished");
[download]

This works in general fine but when there are to much data to archive (around 1.5 GB) I get the error message: "OUT OF MEMORY" and the script stops. When I check the task manager I can see that this happens when the perl process has around 3.2 GB.I did a test with the windows tool 7zip and this works find and the maximum memory this tool is using is around 140MB.

Now my question: What can I do that I do not get this "out of memory"?
Is there a better way to do the archiving to save memory?

Thanks for your help.
de Michi