Dataflow selects random connection after deployment when using Deployment Pipeline

[Anonymous]

Dear PBI community,
I am facing an issue when trying to use Dataflows in Deployment Pipelines. We are using Teradata database as a source.
In DEV phase I configured the dataflow. At the beginning of whole process I chose the server and selected a suitable Gateway connection. Everything works fine in the DEV phase (refreshes, transformations, etc.) 

However when I deploy this Dataflow into the TEST phase I am unable to refresh dataflow and I am not even able to load the tables from the Teradata.

So I tried to check all the transformations steps and I figured out that after the deployment to the TEST Phase there was changed connection to the random one. The listed database containers at the first step - "Source" - are corresponding to the user used in different connection. 

Currently I am not able to use the dataflow at all. My question is pretty straightforward - Did you face the similar issue? Is there any possibility how to "lock" the connection during the deployments? 

I remember that historically there was a issue with dataflows that they were selecting always the first connection used for the Server Connection, however I thought this has been already fixed. 

[Anonymous · Accepted solution]

Hi @Anonymous ,

There are other users who have had similar experiences to you and this is acknowledged as an issue. We have submitted to internal icm 331747353 , engineers are doing their best to fix this issue . please be patient and I will give you feedback here once there is any progress.

Best Regards,
Community Support Team _ Caitlyn

[Anonymous]

Hello, 
do we have any progress regarding to this ICM? Or at least is there any workaround how to work with dataflows? Unfortunately we could not use them once it is not stable. 

Thank you!