Skip to main content

I have a flow automation runs every 5 minutes, and will send out a notification email if a job failed. 

That becomes an issue if a problem that caused the job to fail hasn’t been resolved soon since the automation will continue to send out error email every 5 minutes.

Is there any way to stop sending email after e.g. 10 emails or to reduce the frequency for sending error emails? I still want the automation to run every 5 minutes, even some job failed, but want to get just reasonable amount of error emails. 

Gosh, my first thoughts are to question why the job is failing. Do you know about Rejected Feature handling parameter setting in Form? https://docs.safe.com/fme/2024.2/html/FME-Form-Documentation/FME-Form/Workbench/workspace_parameters.htm#Rejected Once you switch over to Continue Translation, you need to place loggers and/or terminates top capture warning/errors and better inform the transaction log

Refer to the end section of this article https://support.safe.com/hc/en-us/articles/25407430413837-Tutorial-Getting-Started-with-FME-Desktop


Good question J. 

I am relatively new to FME, it’s very likely that my automation has some flaws.  

My automation was originally set up to refresh portal connection weekly so the we connection won’t go expired if unused for two weeks. We recently upgrade GIS portal to 11.3, and heard about this bug https://support.esri.com/en-us/bug/arcgis-enterprise-113-returns-the-too-many-exchange-ref-bug-000172165

so I reconfigured the automation to connect to portal every 5 minutes see if the automation will exhaust the refresh token limits. It hasn’t , but I got lots of error emails when portal server went down unexpectedly. 

I thought FME flow might have some function to get to what I need. Was thinking about logger/counter to track the failed connection attempts. 

I don’t have much experience with rejected feature handling, will look into that.  

Thank you for your suggestion, I will try to get the reject port to a counter, see if I can get what I need.

Thanks.


Now when I start to think about counter, instead of connecting reject port to a counter in the workspace, I may need a different workspace to connect to the job failed port of the previous workspace just for counting, which will then send email if count is certain value...


Reply