I have a model in ArcGIS Desktop that does several Data Reviewer checks. I would like to schedule this model to run weekly using SQL Server. I have exported the model to a python script. It can run successfully from the command line using the ArcGIS python interpreter, like this: c:\python27\arcgis10.6\Python "C:\Scripts\DataReviewer\Scripts\DataReviewerWeeklyChecks.py". When I set up the exact same thing as the only step in a job in SQL Server Agent, however, the step fails with with the message "Executed as user [user]. The step failed." However, it only fails after it appears all the checks have been run, and running data reviewer checks (arcpy.ExecuteReviewerBatchJob_Reviewer) is all the tool does. Furthermore, the DataReviewer output tables (REVCHECKRUNTABLE, RUNBATCHRUNTABLE, REVTABLEMAIN) show that all of the checks are being run successfully (Status = 0), and it's able to find all the same errors it finds when it runs in Desktop. It also takes the same amount of time to run the SQL job as to successfully run the model in Desktop, and only fails at the very end. I have tried using "try" and "except" statements in the Python but the "except" isn't triggered because they Python script isn't registering any errors. The job works fine if I comment out all the Data Reviewer steps and just have it print some text.
Does anyone have any suggestions on how to deal with this? Is there some configuration issue with the SQL server and/or the geodatabase? The Data Reviewer workspace is in a 10.5.1 geodatabase which was recently created. Everything works fine every other way the model is run, including from Desktop, from the command line, from the Python interpreter, and even from Windows Task Scheduler, but we'd really like to use SQL Server if possible.
Edited: I did some additional investigating and using the Windows command line was able to get the error "process exited due to timeout escalation". This happens when the script is done running the Data Reviewer job no matter how long the job takes, even when it's doing a very quick check. What can I do to fix this?
