How to trace each folder, subfolder, datasets, function and SAS programs

[alepage]

Hello,

 I have a SAS programs that call other SAS programs, visit many folders and subfolders, execute many functions and other SAS programs and so on.  Is there a tool availble to follow process flow , either a Unix Script or a SAS one, to better understand a process with no supporting documentation.

I need to get a global idea of this process because we will probably move it to the Cloud Platform in a near future.

I need to understand / know, which SAS functions and programs are essential and which table are read and which one are produced in order to make a good cleanup and may be refined the code with the new tools into the cloud (SAS Viya)

[SASKiwi]

There is really only one tool that can pick apart a complex SAS application to gain a high level of understanding of what it is doing and the outputs it is producing and that is your brain.

There are tools that can parse SAS logs and report on what data is being read and output and what types of SAS processes are being run. PROC SCAPROC can  analyse your SAS logs to identify all inputs and outputs plus the types of processing being done. Your brain needs to look at these results to use your expert knowledge to turn those findings into understanding. 

[alepage]

Hello, I am not familiar at all with the procedure.  Will it be better to add some options such as options mlogic mprint source2 before executing the SAS code to obtain a more detailed log file to use afterwards with proc scaproc .  Is it better to append the 4 logs files produced during the process execution or it is better to scan each file with the proc scaproc?

[SASKiwi]

MPRINT is probably the best macro option to add as you can then see the resolved source code in your logs. If you are doing regular production work I'd recommend running that in batch mode which means you get one SAS log per job.

What you've asked to do is complicated to set up and also to maintain. Personally I don't think it is worth the hassle and I prefer a much lighter touch to administering SAS applications.