Archival question

Question

Wouldn't it be more efficient to compress the collected log file on the server and then transfer the compressed file to the destination. In my case the destination for the compressed files is a different file server. Thus currently the log file is copied to the file server, then the compression routine is run on the log file, in effect copying it right back where it came from and then writing the compressed version back to the server. So the data gets moved a total of 3 times. It just seems more efficient to start a new log file, compress the old one and copy it once in compressed form to the destination and then delete the old log file that was just compressed.

magnum0711 · Answer

Ya I have a similar function going on.  I was more making the suggestion for the deveolpers that's all.  Thanks for the post though I hope it helps people out.

Kuz · Answer

Hi magnum0711,

You're right, this is the normal behaviour if you specify a remote or UNC path for the archive destination.
If this turns out to be a problem for you, can I suggest the following workaround:

1) Change your UNC (or remote) archive destination path to a local path (where exactly doesn't matter, any temporary local location will do, even the default archive destination of "<program files>\syslogd\dated logs")

2) Add an option in the "Archive Options" section of your scheduled archive task to "Run program after all files are moved/copied".  
Specify a file of:   C:\RemoteMove.bat
Specify command-line arguments of: "%FileZipLong" "%FileZipShort.zip"    (including quotes).

3) Create the "RemoteMove.bat" file that will perform the remove move operation, contents as follows:

move %1 \server\share\%2

This procedure will mean that Syslog Server log files will be archived locally, compressed locally, and moved to the remote machine after everything has finished.  I'm also assuming in the scenario above that you are archiving "All files into a single zip", in the archive options.

Hope this helps.