Scheduled Refresh - Intermittent failures due to time out resolved by creating new schedule

[tfgardner]

We have been experiencing an issue with a few of our scheduled refreshes where it will be fine for a period of time with very consistent refresh times and then will fail after the 5 hour timeout (Premium Capacity using an On-PremiseGateway cluster).  A manual refresh will always work.  We have checked server logs and there were no issues with the onsite SQL server that the stored procedure being called lives on.  Sometimes the next day it will just start working again but other times we have to remove the schedule and start a new one in order to get it consistently working again and it will be fine for another few weeks and then the problem starts again.  This has happened several times on a few different data sets.  Is there a known issue with this or is there somewhere we can check on our gateway setups to identify why the scheduled job will go into this failure cycle?

[v-piga-msft]

Hi @tfgardner,

By my research, this used to be an issue several days ago. Some customers in North Europe and North Central US may experience intermittent refresh issues during peak usage times. 

Currently, this issue has been fixed on 2019/1/25. Please have a try again.

If you still have problem, please let me know.

Best Regards,

Cherry

[tfgardner]

Looks like the schedule ran fine on 1/26 and 1/27 - but then went into a spin again this morning and failed after 5 hours.

 Will run a manual and see if tomorrow am works and if not will delete schedule and start a new one.