Using Flow

[YannRC]

I really want to push the topic of flows to the maximum (and develop the low‑code/no‑code approach). I’m already quite comfortable with flows, but I want to take things to the next level. I know there’s an issue with promotes that should be fixed in the next LTS release. I’m not sure if what I’m seeing is related to that.

My goal is the following: I take a detailed source table, aggregate it to produce a summary, and archive the daily results. My flow works perfectly when I run it manually, but it fails when it’s scheduled.

Here’s what I’ve done: I start from a detail table, run a query to aggregate it, and save the result in a CAS library. In that same library, I previously created an archive table with the same structure. Then I add a data step to insert rows. Finally, I add another query to work around the promote issue.

I believe everything is set up correctly. But when it runs automatically, it seems to start from scratch every time, as if the table generated by the row‑insertion step was never promoted.

Can you help me?

[gustaft · Accepted solution]

Hi YannRC, I would start by examining the log from the failed scheduled flow and compare it with the log from a successful manual run. 

Maybe there is some difference in the Compute and Batch contexts that can explain why it fails when scheduled? 

If you cannot figure out yourself why it fails I would recommend you contact SAS Technical Support, https://support.sas.com/en/technical-support.html

Open a Support case from the SAS Customer Service Portal or send an email to support@sas.com.

[YannRC]

Thanks, i open a case because the 2 log are similar

[Anacreo]

I wanted to add that I just resolved an issue similar to this for a user, and can't tell you how valuable Visual Studio Code is in this situation.

If you ctrl-click on two open file tabs and then right click on the tab, there is a "compare" option.  It's actually a very good comparer if you ask me.

[YannRC]

Thanks