Data Refresh timeout

[Ander]

Scheduled Data refresh in pro workspace takes more than 2 hours leading to timeout (used to take only up to 20 minutes). No changes were made at all to the report and underlying data model/power query. Data refresh has failed since June 7 and is only alleviated when reducing the amount of data queried by a drastic amount (pulling 2 years of data vs 5 years). This hasn't been an issue in the past so I am confused why we are seeing issues now and I wonder why it happened very suddenly. This impacts both dataflows in pro workspace and regular published datasets. This only appears to occur with reports that connected to oracle via on-premises gateway.

[Anonymous]

Hi @Ander ,

We have a internal icm 314528591 with same experiences with you, The engineer's recommendation is.
Check the gateway performance logs, make sure your gateway CPU usage is not too high.

Please advise the customer to scale out their cluster by adding another cluster member. If they do not do so, they may see crashes of the gateway process or failed refreshes since we are hitting the memory limit of the system.
https://docs.microsoft.com/en-us/data-integration/gateway/plan-scale-maintain

Manage on-premises data gateway high-availability clusters and load balancing | Microsoft Docs

Best Regards,
Community Support Team _ Caitlyn