Refresh Failures | PostgreSQL: Exception while reading from stream

[harisfatik1]

Hi,

We have been facing this issue for quite some time. The dashboards failed during the scheduled refresh and returns the following error:

[ibarrau]

Hi. It looks like there is an issue between the source and service or gateway (depends on your connection). I'm not sure about the fix, but I have found this on the forum. I hope it helps you https://community.powerbi.com/t5/Desktop/Postgresql-Exception-while-reading-from-stream/m-p/231070

Regards

[harisfatik1]

We have tried this fix but still, the dataset refresh returns the same error.

[jaap_olsthoorn]

Hey @harisfatik1 , did you ever work this out? I'm struggling with the same error. Just started happening suddenly after being fine for literal years...

[hfatik]

Hey, it took us a couple of weeks to investigate the root cause of this error. There was a limitation on our Greenplum Database to accommodate below:

1. No. of Total Connections: It can only handle a max of 350 connections at a particular time.
2. No. of Queries Running in Parallel: It was configured to handle only 30 queries to run in parallel for the assigned DB user group (for PowerBI refreshes).

How did we resolve this?

We had to modify the configuration of the Greenplum database and increased the Concurrent Queries limit (from 30 to 200) as we can see PowerBI tried to execute 154 queries just for one dataset (Star Schema with 18 tables).

What this means is that PowerBI is trying to establish new connections to execute queries on the Greenplum database but they are going straight to “Queued” status and hence the initial handshake acknowledgement between Source (PowerBI) and Destination (Greenplum Database) doesn’t happen and the connection gets aborted by PowerBI after waiting for a certain duration.

[jaap_olsthoorn]

Thanks a lot for the response! Unfortunately, my issue was a lot less complicated. A colleague of mine told me that indexes persist when you delete and recreate a table, but we learned the hard way that that was not correct. So all that was wrong on mine was that we accidentally deleted an index and it made our queries so slow to the point of timing out. 😞

Hopefully your response helpt someone in the future though! Much appreciated!