Error when updating Power BI connected to Impala . Gateway partially connected.

[Ana67]

 I have a PowerBi application with some dashboards. The dashboards are updated according to the configuration at least 2 times a day. The dashboards connected to Impala sometimes do not update and I get this error (see below). It does not happen all the time but many times. I have to update it manually every time I get the error. 

I use cloudera ODBC driver (64 bit) and basic gateway (PowerBi Cloud): on-premise data gateway.

If a check the gateway, I get the message that the "Local gateway partially connected: The test connection was partially made because the gateway cluster is accessible, but not all members are online. Open the gateway cluster details pane to determine which gateway members are inaccessible."

Can you help me to solve the update problems?

Thank you,

Error message:  

• {"error":{"code":"DM_GWPipeline_Gateway_MashupDataAccessError","pbi.error":{"code":"DM_GWPipeline_Gateway_MashupDataAccessError","parameters":{},"details":[{"code":"DM_ErrorDetailNameCode_UnderlyingErrorCode","detail":{"type":1,"value":"-2147467259"}}, {"code":"DM_ErrorDetailNameCode_UnderlyingErrorMessage","detail":{"type":1,"value":"ODBC: ERROR [HY000] [Microsoft][DriverSupport] (1170) Unexpected response received from server. Please ensure the server host and port specified for the connection are correct."}},{"code":"DM_ErrorDetailNameCode_UnderlyingHResult","detail":{"type":1,"value":"-2147467259"}},{"code":"Microsoft.Data.Mashup.ErrorCode","detail":{"type":1,"value":"10478"}},{"code":"Microsoft.Data.Mashup.ValueError.DataSourceKind","detail":{"type":1,"value":"Impala"}},{"code":"Microsoft.Data.Mashup.ValueError.DataSourcePath","detail":{"type":1,"value":"md-iplb-rbap.emea.bosch.com:21050"}},{"code":"Microsoft.Data.Mashup.ValueError.OdbcErrors","detail":{"type":1,"value":"#table({\"SQLState\", \"NativeError\", \"Message\"}, {})"}},{"code":"Microsoft.Data.Mashup.ValueError.Reason","detail":{"type":1,"value":"DataSource.Error"}}],"exceptionCulprit":1}}} Table: qmm_process_uss5_fermasilb_batch.URI de clústerWABI-NORTH-EUROPE-L-PRIMARY-redirect.analysis.windows.net

[v-pnaroju-msft]

Hi Ana67,

We are following up once again regarding your query. Kindly confirm whether the issue has been resolved through the support ticket with Microsoft. If the issue has been resolved, we would appreciate it if you could share the resolution or key insights here to assist others in the community.
In case we do not receive a response, we shall proceed to close this thread.Should you require any further assistance in the future, we encourage you to reach out via the Microsoft Fabric Community Forum and initiate a new thread. We will be happy to assist you.

Thank you.

[v-pnaroju-msft]

Hi Ana67,

We are following up to check whether you have raised the support ticket. If you have already done so, we kindly request you to share your feedback on the issue raised.
If a solution has been provided, we would appreciate it if you could share it with the community and mark it as the accepted solution. This will help others facing similar challenges and benefit the broader community.

Should you have any further queries, kindly feel free to contact the Microsoft Fabric community.

Thank you.

[v-pnaroju-msft]

Hi Ana67,

Thank you for your update, and we sincerely apologise for the inconvenience caused.

Since the issue remains unresolved, we kindly request you to raise a Microsoft support ticket using the link provided below. The support team will be able to offer specific insights related to your account and suggest appropriate resolutions.
Microsoft Fabric Support and Status | Microsoft Fabric

If you find our response helpful, we would be grateful if you could mark it as the accepted solution. This will assist other community members who may be facing similar queries. Should you have any further questions, please do not hesitate to reach out to the Microsoft Fabric community.

Thank you.

[v-pnaroju-msft]

Hi Ana67,

We wanted to check in regarding your query, as we have not heard back from you. If you have resolved the issue, sharing the solution with the community would be greatly appreciated and could help others encountering similar challenges.

If you found our response useful, kindly mark it as the accepted solution to guide other members.

Thank you.

[Ana67]

Hello Comunity, 

I have tried all the recomendations /ideas but I 'm still having the intermittent refresh failures.

I'm still looking for the solution.

I will post it as soon as I find it.

In any case, I appreciate all your recomendations,

Thank you, 

[v-pnaroju-msft]

Hi Ana67,

We have not received a response from you regarding the query and were following up to check if you have found a resolution. If you have identified a solution, we kindly request you to share it with the community, as it may be helpful to others facing a similar issue.

If you find the response helpful, please mark it as the accepted solution, as this will help other members with similar queries.

Thank you.

[v-pnaroju-msft]

Thankyou, @nilendraFabric, @BhavinVyas3003, and @rohit1991, for your responses.

Hi Ana67,

We sincerely appreciate your inquiry through the Microsoft Fabric Community Forum.

Based on our understanding, the intermittent refresh failures may be due to the following reasons:

1. Partial availability of the gateway cluster, where some nodes are offline.
2. An unstable ODBC layer with Impala, resulting in network handshake or timeout errors (ODBC: ERROR [HY000]).
3. The gateway service account lacking the necessary network or certificate permissions during unattended scheduled refreshes.

Please kindly follow the additional steps outlined below, which may help in resolving the issue:

1. The Windows service account running the gateway (PBIEgwService) must have Read access to any SSL certificates used by the Impala DSN, as well as network permissions to reach the Impala server DNS and port. If a custom service account is used, please ensure it is not restricted by Group Policy or firewall settings.
2. Enable detailed logging in the gateway to capture any transient network or certificate issues during service refreshes. You can do this by modifying the file located at:
   C:\Program Files\On-premises data gateway\Microsoft.PowerBI.DataMovement.Pipeline.GatewayCore.dll.config
   Set the logging level to Verbose, then reproduce the error and collect the logs for further analysis.
3. Kindly avoid hosting the gateway on machines that also run user applications or have antivirus/firewall software which may interfere with ODBC communication during idle or scheduled hours. It is advisable to use a dedicated VM or server with a static IP and always-on connectivity.
4. The message “Unexpected response from server” could also be caused by Impala’s backend limits such as query queue, memory, or user session caps. Please coordinate with your Cloudera administrator to review the logs for query failures or connection drops.

Additionally, please refer to the following link for further guidance:
Monitor and optimize on-premises data gateway performance | Microsoft Learn

If you find our response helpful, we kindly request you to mark it as the accepted solution and provide kudos. This will assist other community members who might face similar issues.

Should you have any further queries, please feel free to reach out to the Microsoft Fabric community.

Thank you.

[rohit1991]

Hi @Ana67 ,

This type of intermittent refresh failure with Power BI and Impala through the on-premises gateway is a pretty common headache, but it’s usually fixable once you check a few core areas. Here’s a practical rundown based on your scenario and what’s worked for me and others:

1. Double-Check Gateway Cluster Health: Go to Power BI Service → Manage Gateways. Make sure all gateway nodes are online, healthy, and up to date. Even one offline/lagging node can cause “partially connected” errors and make refreshes unreliable. If any node is down, either bring it back online or remove it from the cluster.

2. ODBC Driver Version & Connection Upgrade to the latest Cloudera ODBC driver for your version of Impala. Old drivers often cause handshake and timeout issues, especially after an update. On the gateway machine, use ODBC Data Source Administrator or DBeaver to test the connection to your Impala server. Make sure the host, port, and credentials work outside of Power BI.

3. Network & Permissions: Check for any network/firewall rules that might block ODBC or SSL traffic between the gateway and Impala server, especially if the error is intermittent. The gateway service account needs full permissions to reach the Impala server and any SSL certs if you’re using encrypted connections.

4. Increase ODBC Timeout and Enable Retry: In your ODBC DSN settings, set Socket Timeout to at least 60 seconds (sometimes more is needed). Enable any available retry options in the DSN or gateway settings.

5. Stagger Power BI Refreshes: If you’re running multiple scheduled refreshes close together, try spacing them out by a few minutes to avoid resource/contention issues.

6. Use Power Automate or REST API for Retry Logic: If manual refreshes work but scheduled ones sometimes fail, consider using Power Automate or the Power BI REST API to automatically retry failed refreshes. This can cut down on manual intervention.

[BhavinVyas3003]

Hi @Ana67 ,

Fix Gateway Cluster:

• Go to Power BI Service > Manage Gateways.
• Ensure all gateway members are online and updated to the latest version.
• Remove or fix any offline/inaccessible nodes.

Test and Fix ODBC Connection:

• Use ODBC Data Source Admin or tools like DBeaver on the gateway machine.
• Verify connection to: md-iplb-rbap.emea.bosch.com:21050.
• Check network/firewall/DNS if connection fails.

Use Compatible Cloudera ODBC Driver:

• Confirm driver version is compatible with your Cloudera Impala version.
• Download/update from Cloudera official site.

Increase Timeout and Enable Retry:

• In the ODBC DSN settings, increase Socket Timeout to at least 60 seconds.
• Enable retry options, if available.

Stagger Power BI Scheduled Refreshes:

• Space out refreshes to avoid overlapping and reduce load.

Use API or Power Automate for Backup Refresh:

• Set up Power BI REST API or Power Automate to retry refreshes on failure.

[nilendraFabric]

Hi @Ana67 

The fact that manual refreshes work suggests the core connection is valid, but scheduled/service refreshes via the gateway hit issues—often because the gateway service environment differs from your desktop environment

Open the gateway cluster details in Power BI Service and bring all members online.

Upgrade to the latest Cloudera ODBC driver for Impala.

If SSL is enabled, ensure certificates are valid and accessible to the gateway service