Report refresh failure when published in new workspace

[joshishweta]

I had 2 different reports with separate datasets published in the same Workspace say X", both working fine no issue with refresh, but to organize the workpces by department. I created two new workspaces as X1 & X2, then published i.e. moved each report in appreciate workspace, set the scheduled refresh at 6:00 AM as per our IT dept. For both the reports option to run in parallel is set to off (unchecked in the setting for each report). 

So only change here is new workspaces, nothing was changed with the report queries, still refresh is failing for both reports.

My question is when option to "run in parallel is turned off for the report dataset", is it limited to reports in the same Workspace or across all workspaces with same owner?

[lbendlin]

What you referring to is the parallel load of tables from your dataset's data sources.  Say you have one dataset with five SQL server tables, each pretty large and maybe even with complex filters.

If these tables all reside on the same server then yes, preventing parallel load WILL help. But if they reside on different SQL servers then it doesn't really matter.

In you refresh failure error message you should find the reason for the failure. It could be a simple timeout, or it could be that one of the requests has been selected as the table lock victim if both datasets point to the same table, or if your data retrieval coincides with data updates etc.

Post the refresh error message for BOTH datasets.

[joshishweta]

Thank you @lbendlin for your quick response. Even though error shows "it ran out of memory trying to run the query" but, if I publish the reports manually it takes less than a minute to publish in power bi service. 

All Reports have the same datasource Vertica (ODBC connection), here is the error message - 

Dataset 1 - 

Dataset2- 

[joshishweta]

Dataset2 -

[lbendlin · Accepted solution]

These messages point to a culprit outside of Power BI's control. Talk to your Vertica DBAs.

[joshishweta]

Thank you @lbendlin. I'll do that. I really appreciate your help, thank you!