Error in the Service with Refresh- DM_GWPipeline_Gateway_MashupDataAccessError

[aizhan_otp]

Hi everyone!

Can you tell maybe someone has encountered such an error. Each time the error refers to different tables.

There is only one gateway connection configured.
Some nuances: the connection as a file source is through the ODBC connector (system DNS with drivers to PostgreSQL).The database is hosted in Amazon RDS.

In the desktop version, the file takes a very long time to update, about 5 hours. But when schedule refresh wants to update, it doesn't work at all. Until 29thMay  schedule refresh works consistently and processing time were over 2 hours

Please someone tell me an answes or what is the solution to this problem

[aizhan_otp]

Do you mean rewriting the query script on dowloanding tables?
By the way my tables are some agregation/ 
Recently average complete refhresh  were 2 hours

[lbendlin]

Sounds like you already found your answers yourself.

[lbendlin]

You need to make your query run faster.  The 5 hour timeout is a hard limit in the service for semantic models.  There is no such limit for dataflows, but you don't want to go that route as the dataflow will incure a gigantic CU cost upon completion or failure.

As per Shannon/Nyquist your semantic models should complete refreshes on average in 2.5 hrs.