Overview:

This technote will describe how to add additional processors to increase Automated File Processing for ProjectWise.  Please take note of the minimum requirements above, this will not work properly and is not supported unless the minimum requirements are met.

Please make a backup of your Orchestration Framework database before following this technote.  This will allow you to scale back to the original configuration in case this does not work out for you.  If it does not work for you, just restore the original database over the newly configured database.  

In order to get the processing to work as fast as possible, please follow this blog on how to properly configure the Full Text Indexing values in your ProjectWise Administrator:

http://communities.bentley.com/products/projectwise/content_management/b/weblog/archive/2012/08/08/optimizing-automated-file-processing.aspx

The AFP-multiple-processors.zip file can be found here.   There are instructions as well as notes inside of the file:

ProjectWise CE and later:

Due to some code changes the below scripts will no longer work if you are running CONNECT Edition, so we have ported the changes into PWPS_DAB to make things a little easier for the users.  I always recommend starting with a fresh Orchestration Framework Database but if you are upgrading your Orchestration Framework database then please run Update-PWAFPQueueNames to update your queue names to the appropriate naming convention.  If you are starting with a fresh Orchestration Framework database please use the Copy-PWAFPProcessors cmdlet and follow the steps below.  This will also only add in another row of processors as we have observed one additional path is fast enough for the majority of our users.

ProjectWise v8i:

http://communities.bentley.com/products/projectwise/content_management/m/mediagallery/269773.aspx

Setup post CE:

1. Stop Orchestration Framework on your head and tail server.
2. Run the Copy-PWAFPProcessors using the name of your indexing server in the -MachineName parameter.
3. Start your Orchestration Framework Administrator
4. Start Orchestration Framework on your head and tail
5. Manually map the new processors by dragging the nodes from the left pane into the editor window in the Orchestration Framework Administrator to represent the picture below. This is an example of FTR.
6. After manually mapping the processing should resume.

Setup prior to CE:

1. Stop Orchestration Framework on your head and tail server
2. Run this query on your Orchestration Framework database to check how many processors you have and to make sure the server name is right: select * from orchShepherdProcesses
3. Edit the @MachineName value in one of the attached sql scripts to represent your Tail server
4. Run whichever sql script you would like to use after editing the @MachineName value
5. Start your Orchestration Framework Administrator
6. Start Orchestration Framework on your head and tail
7. Manually map the new processors by dragging the nodes from the left pane into the editor window in the Orchestration Framework Administrator to represent the picture below. This is an example of FTR.
8. After manually mapping the processing should resume.

Notes:

After testing it appears that adding one extra set of processors to bring it up to two sets of processors is the best approach, three total sets have been attempted on some servers and some work but some it tended to max the server's resources out and make it unresponsive.  It is your choice if you would like to go above 2, but be warned you will not see much benefit.  Also when upgrading you will want to start with a fresh Orchestration Framework database since this does not convert over when doing the upgrade, after your upgrade please feel free to run the scripts/cmdlets again.  Please let TSG know if you have any questions on this technote.