Error interacting with REST API: Timeout was reached

[Anonymous]

Hi,

I have a working dataset in service which is around 4GB with 18 months of data which runs without any issues. The source is BigQuery and I am running import mode on all tables. The incremental refreshes are running smoothly and it's located in a P2. 

Now, a new view was created that I connected to with the already existing model and I receive the following error when doing a refresh in the service:

• Data source errorODBC: ERROR [HY000] [Microsoft][DriverOAuthSupport] (8701) Error interacting with REST API: Timeout was reached ERROR [HY000] [Microsoft][DriverOAuthSupport] (8701) Error interacting with REST API: Timeout was reached. The exception was raised by the IDbCommand interface.

Before I had an issue with the REST API so i added the UseStorageAPI=false which made it work before. I also have quite high permissions in BigQuery which I am able to query.

Anyone has any other experiences on this front?

Kind regards,

Jari

[siarejfm]

Hi @Anonymous 

I am sorry to hijack this thread for another topic, but it just striked me that you seem to have managed to setup incremental refresh with BigQuery as a datasource?

We have an open ticket with MS Support about this in particular, and was informed that this is currently not supported by MS. The functionality hasn't been made unavailable for that source, but it seems like the timestamps sent in the jobs created by PowerBI and sent to BigQuery is formated wrongly.

Would it be possible to share thoughts about this with you, since it seems that you have managed to overcome this issue?

Thanks for letting me know.

UPDATED:
-----------
Link to idead thread created in regards to above:
https://ideas.powerbi.com/ideas/idea/?ideaid=aa275090-5641-ec11-a3ee-501ac50a6ef7

Regards Simon

[Anonymous]

Hi @Anonymous ,

Could you tell me if your problem has been solved? If it is, kindly Accept it as the solution. More people will benefit from it.

Best Regards,
Eyelyn Qin

[Anonymous]

Hi,

We will move it to a P3, seems like the P2 was way too full all the time with the different workspaces. Assuming that the Model Refresh Parallelism in P2 refers to all workspaces and datasets in the capacity?

Kind regards,

Jari

[Anonymous · Accepted solution]

Hi @Anonymous ,

The.pbix files represent data in a highly compressed state. The data will likely expand when loaded in memory, and from there it may expand several more times during data refresh.

And Scheduled refresh of large datasets can take a long time and be resource-intensive. It's important to not schedule too many overlapping refreshes since the maximum Model Refresh Parallelism in P2 is 12.

Please check the memory left and the refresh progress.

Refer to:

Connect to a Google BigQuery database in Power BI Desktop - Power BI | Microsoft Docs

Solved: Data Refresh Timeout Limit for P1, P2 and P3? Time... - Microsoft Power BI Community

Solved: Dataset Refresh Failing due to Timeout - Microsoft Power BI Community

How to configure workloads in Power BI Premium - Power BI | Microsoft Docs

Best Regards,
Eyelyn Qin
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.