Before the data import for finished, its data source timed out. Double-check whether that data sour

[VladT]

Hello,

We have a couple of datasets that are constantly failing with the Before the data import for finished, its data source timed out. Double-check whether that data source can process import queries, and if it can, try again. error. 

The only remedy is to restart the gateway, any assistance would be appreciated.

Cluster URI:WABI-US-NORTH-CENTRAL-B-redirect.analysis.windows.net

Activity ID:a6a96350-a1aa-4303-af23-81b87bd2cc92

Request ID:30956802-f027-44bd-96cd-08b8d9cf4cc2

Time:2022-09-18 23:07:50Z

[v-cazheng-msft]

Hi @VladT,

May I know what’s the data source of your report and what’s the connection mode? What’s the version of your gateway? Is it the latest version? Did you install your gateway as the same machine as the data source?

Best Regards,

Community Support Team _ Caiyun

[VladT]

Hi Caiyun,

Please find the responses below.

• What’s the data source of your report, and what’s the connection mode?  Several SQL Views, some traversing multiple tables on various SQL servers. 
• What’s the version of your gateway? Is it the latest version? August, which I believe is the latest:
  
• Did you install your gateway on the same machine as the data source? No, we have dedicated gateway servers clustered for redundancy.

Best regards

Vladimir

[VladT]

We are seeing this issue in another data sources as well here is the error message:

Processing error: Your data gateway is offline or could not be reached. Ensure the computer running the gateway is on during the refresh period.
Cluster URI: WABI-US-NORTH-CENTRAL-B-redirect.analysis.windows.net
Activity ID: 3bba1ce1-2d55-4ae4-ba9e-527e4351d64b
Request ID: 31e69ee8-c9b8-4081-91d0-fb1fdd58a4cf
Time: 2022-09-29 21:12:29Z

[edwsmeadmin]

@v-cazheng-msft

Last refresh failed: 10/11/2022, 12:37:58 PM
Before the data import for finished, its data source timed out. Double-check whether that data source can process import queries, and if it can, try again.Hide details

Cluster URI:	WABI-INDIA-CENTRAL-A-PRIMARY-redirect.analysis.windows.net
Activity ID:	09bc231d-6726-4fbe-a12f-35282637326a
Request ID:	1db755dd-9681-488e-9b34-1c85f73d994e
Time:	2022-10-11 07:07:58Z

[knovy99]

@v-cazheng-msft hoping for guidance.  We see this issue often on our SQL sources.

[SA-ERUY-ETL]

same error here, did you find a solution?

[kumar27]

@VladT , i am getting the same error , were you able to find the solution ?Any help will be appreciated .

many thaks 

[Lun]

@kumar27 

Try restarting the machine that has the gateway installed on.
Yesterday our datasets went out and this morning, after restarting machine with gateway and turning scheduled refresh back on it seems to be working again.
Both on demand refreshes and scheduled ones have cleared for us so far.

We are running gateway in personal mode.
Although the gateway reads an "OK" on dataset settings page there seems to be some issue with it.
Over the last few weeks about 5 of our largest data sets started to become inconsistent with scheduled refreshes and eventually went down completely.

At first it impacted just some of the scheduled updates and error message was:
"Failure details: The last refresh attempt failed because of an internal service error. This is usually a transient issue. If you try again later and still see this message, contact support."

Eventually the scheduled refresh stopped completely for all datasets and error message switched out to:
"Before the data import for finished, its data source timed out. Double-check whether that data source can process import queries" which in turn also disabled the scheduled refreshing in general.

Only thing I can think of that might have caused this, might be general growth of one of our datasets.
Maybe it has gotten so large that it requires a lot of time to process and somehow causes other updates using the same gateway and accessing the same server to get bunched up somehow.
Admittedly we are running a rather old installation of gateway and might update in future if problem persists, but it had no issues like this for years so we havent bothered to update it.