I have an enterprise data gateway which resides on the same server as my SQL Server data source. I receive the "Your gateway is offline or cannot be reached" error message intermittently during daily refreshes.
When this occurs I log onto the server to ensure the gateway is up and running. The gateway will show up as "online and ready to be used" and the network port test succeeds. Yet when I try to run a refresh on the dataset in Power BI service this issue remains. I have tried running a trace on the data source while attempting to refresh the dataset but there doesn't seem to be any connectivity attemts to the datasource coming from the Power BI user.
The gateway logs are showing one of the following errors:
1) A network-related or instance-specific error occurred while establishing a connection to SQL Server. The server was not found or was not accessible. Verify that the instance name is correct and that SQL Server is configured to allow remote connections.
2) The credentials provided for the SQL source are invalid.
3) Microsoft.PowerBI.DataMovement.Pipeline.Diagnostics.SqlDataAccessException: SqlException encountered while accessing the target data source
4) Microsoft.PowerBI.DataMovement.Pipeline.Diagnostics.MashupDataAccessCredentialException: Missing or invalid credentials for data source. Reason = 'AccessUnauthorized', DataSource = '<pi>XXXXX</pi>', DataSourceReference = '', DataSourceOrigin = '', DataSourceReferenceOrigin = ''.; ErrorShortName: MashupDataAccessCredentialException[ErrorCode=-2147467259,HResult=-2147467259]/Wrapped(MashupCredentialException)[ErrorCode=-2147467259,HResult=-2147467259]
5) Microsoft.PowerBI.DataMovement.Pipeline.Diagnostics.GatewayCancellationException: The operation failed due to an explicit cancellation.
On occasion I find that there are 2 Power BI datasets that refresh at the same time for the same source data. One will refresh fine while the other will give the "gateway offline or can't be reached" error.
The issue is ongoing since October. I updated the gateway to the latest version in early December (November version) but the issue is ongoing.
Does anyone have any ideas why this is happening or a possible solution?