We often see Anaplan Connect scripts created ad-hoc, as new actions are added, or updating existing scripts with these new actions. This works when there is a limited number of imports/exports/processes/etc. running, and when these actions are relatively quick. However, as a models and actions scale up and grow in complexity, this solution can become very inefficient. Either scheduling dozens of scripts, or trying to manage large, difficult to read scripts.
I prefer to design for scale from the outset. My solution utilizes batch scripts that call the relavant Anaplan Connect script, passing the action to run as a variable. There are a couple ways I've accomplished this: dedicate a script to execute processes and pass in the process name, or pass in the action type (-action, -export, etc.) and name as the variable. I generally prefer the first approach, but you want to be careful when creating your process that it doesn't become so large that it impacts model performance. Usually, I will create a single script to perform all file uploads to a model, then run the processes. In my implementations, I've written each Anaplan Connect script to be model specific, but you could pass the model ID as a variable as well.
To achieve this, I create a "controller" script, that calls the Anaplan Connect script, which would look something like this:
@echo off for /F "tokens=* delims=" %%A in (Demand-Daily-Processes.txt) do ( call "Demand - Daily.bat" %%A & TIMEOUT 300) pause
This reads from a file called Demand-Daily-Processes.text, reads a line which contains the name of the process as it appears in Anaplan, e.g.,
Load Master Data from Data Hub ... Load Transactional Data from Data Hub
Then it calls the Anaplan Connect, passing this name as a variable. Once the script completes, the controller waits 300 seconds before reading the next line and calling the AC script again. This timeout is there to give the model time to recover after running the process and prevent any potential issues executing subsequent processes.
The Anaplan Connect script itself looks mostly as it does, except in place of the process name, we use a variable reference.
@echo off set AnaplanUser="" set WorkspaceId="" set ModelId="" set timestamp=%date:~7,2%_%date:~3,3%_%date:~10,4%_%time:~0,2%_%time:~3,2% set Operation==-certificate "path\certificate.cer" -process "%~1" -execute -output "C:\AnaplanConnectErrors\<Model Name>-%~1-%timestamp%" rem *** End of settings - Do not edit below this line *** setlocal enableextensions enabledelayedexpansion || exit /b 1 cd %~dp0 if not %AnaplanUser% == "" set Credentials=-user %AnaplanUser% set Command=.\AnaplanClient.bat %Credentials% -workspace %WorkspaceId% -model %ModelId% %Operation% @echo %Command% cmd /c %Command% pause
You can see that in place of declaring a process name, the script uses %~1. This tells the script to use the value of the first parameter provided. You can set to 9 variables this way, allowing you to pass in workspace and model IDs as well. This also creates a timestamp variable with the current system time when executed, then uses that and the process name to create a clearly labled folder for error dumps. eg. "C:\AnaplanConnectErrors\Demand Planning-Load Master Data from Data Hub-dd/mm/yyyy time".
By using this solution, as you add processes to your model, you can simply add them to the text file (keeping them in the order you want them executed), rather than editing or creating batch scripts. Additionally, you need only schedule your controller script(s), making maintenance easier still.