Description
-----------------
To keep replica vaults up to date with the latest versions from the master
storage cluster, folders should be replicated on a regular basis. The interval
of replication should be determined by the administrator based on the rate at
which data is added to the master vault. 

There are two methods for replicating data:
 -- Pro/ADMIN folder tab
 -- <proiclient_loadpoint>/bin/replicate_folder script

This document discusses techniques that may be implemented with the
replicate_folder script to maintain up-to-date replica storage clusters in an
automated, "hands-off" fashion. Additional information regarding the
replicate_folder script can be found at:

http://www.ptc.com/cs/tan/102718.htm

Online Help Topics (available as part of the proiclient):
 -- Updating Storage Clusters using replicate_folder
 -- Purging Replicated Storage Clusters

Resolution
-----------------
The best choice for automating the replication process is to use the command
line script, replicate_folder. This script may be called from a master script
to replicate many Commonspace folders. The master script may then be scheduled
to execute as a UNIX cron or Windows AT job during a period of low network
activity. Below (between the * characters) is an example of a master script
designed to replicated two Commonspace folders:

************
@echo off

call <path_to_proiclient>\bin\replicate_folder.bat INTRALINK INTRALINK "/Root Folder/Products/Project X" -D -f
call <path_to_proiclient>\bin\replicate_folder.bat INTRALINK INTRALINK "/Root Folder/Products/Project X" -R

call <path_to_proiclient>\bin\replicate_folder.bat INTRALINK INTRALINK "/Root Folder/Products/Project A_0_2002" -D -f
call <path_to_proiclient>\bin\replicate_folder.bat INTRALINK INTRALINK "/Root Folder/Products/Project A_0_2002" -R
************

Notice that there are two lines specified for each Commonspace folder. The first
line executes replicate_folder with -D, while the second line executes the -R
option. As the documentation states the -D option will delete objects from the
replica cluster that have been deleted from the master storage cluster.
Executing this as part of the regular replication routine is a good maintenance
practice to observe. Obviously, there is no point in preserving objects in the
replica storage cluster that have been deleted from the master cluster. There is
no metadata that corresponds to the replica data and by removing it, no hard disk
resources are wasted at the replica site. The master script will then execute the
-R option to replicate any new master data. 

To gain a high level of confidence in the script, TEST IT thoroughly. Once the
script has been validated, schedule it to run at a specified interval. The folder
replication process is fully automated at this point.
