Scheduled refresh randomly failing every day with error: "DM_GWPipeline_Gateway_CanceledError".

[ezaidi]

Hi, for years we have been using the same gateway without problems and many reports would refresh just fine on schedule without issues. Since around January 11th 2025, every morning almost at random, different reports fail to refresh with the same error.

{"error":{"code":"DM_GWPipeline_Gateway_CanceledError","pbi.error":{"code":"DM_GWPipeline_Gateway_CanceledError","parameters":{},"details":[],"exceptionCulprit":1}}} Table: "Some random table".

We have about 20 different reports and maybe less than half of them fail with this error message and its not always the same report every day. Some days they refresh and other days they wont. I have to do an on demand refresh and it works.

This is really frustrating and I dont know why this is happening now after all these years. The on-prem data gateway shows that our gateway version is 3000.142.14 (September 2022)

Is anyone else experiencing this?? What can we do to fix it? Do I need to update it?

Edit: Please note that I had SHIR installed on the same server and have since disabled the service from running. This has still not helped Onpremise Data Gateway. Bi-service refresh is still failing in the morning on many reports. I have not updated data gateway version yet, but as state by someone else below, even updating has not helped.

[v-saisrao-msft]

Hi @ezaidi,
Could you please confirm if the issue was resolved after raising a support case? If so, we’d appreciate it if you could share the solution to help others in the community. As we haven’t heard back, we’re closing this thread. For any further issues, please raise a new thread in the Microsoft Fabric Community Forum — we’ll be happy to assist.

Thank you for being part of the Microsoft Fabric Community.

[v-saisrao-msft · Accepted solution]

Hi @ezaidi 

Thank you for reaching out to the Microsoft Forum community.

Apologies for the delayed response. Could you please confirm if the issue has been resolved?

If the issue persists, kindly raise a support ticket with Microsoft Fabric Support. This will enable the support team to provide specific guidance and assistance.

To raise a support ticket for Fabric and Power BI, kindly follow the steps outlined in the following guide:

How to create a Fabric and Power BI Support ticket - Power BI | Microsoft Learn

Thank you.

[mcaraballo]

Hello, Any updates on this?

[ezaidi]

Im still having the same issue. I have not found a resolution to this yet.

[v-saisrao-msft]

Hi @ezaidi,

I wanted to follow up on the "DM_GWPipeline_Gateway_CanceledError" issue. Has your issue been resolved or were you able to raise a support ticket with Microsoft Fabric Support?

Your feedback would be greatly appreciated, as it can help us better understand and address similar issues in the community.

Thank you!

[AlexisAllerborn]

HI, We're also experiencing the same issue on our shared capacity since Jan the 13th. 

We do have tired installing the latest OPDG version but it didn't fix it.

[ezaidi]

At first I thought that there was maybe a conflict with SHIR being also installed on that server for ADF but I have since disabled the service so that OPDG doesnt conflict because I read somewhere that there may be a conflict with SHIR and OPDG running on the same system. So far this has not helped at all. So the latest version did not help either? This has to be a BI Service issue. We may have to escalate this with MS directly.

[rohit1991]

Hi, @ezaidi 

“DM_GWPipeline_Gateway_CancelledError” issue on scheduled refreshes. I know a lot of you have tried the basics, so here’s a checklist that sometimes helps uncover what’s really going wrong:

1. Update the Gateway: Double-check that you’re on the absolute latest gateway version and reboot the server after updating even a small patch can make a difference.

2. Monitor Server Resources: While refreshes are running, keep an eye on CPU, RAM, disk, and network usage. Spikes or sustained high load (especially during peak hours) can trigger these errors.

3. Review Gateway Logs: Right after a failure, grab the gateway logs and look for error patterns or anything that stands out around the time of the crash. Correlation IDs from Power BI Service refresh history are super useful if you need to escalate.

4. Spread Out Refresh Schedules: If multiple reports are scheduled close together, try to stagger them. Overlapping refreshes can overload the gateway or backend sources, especially if they’re large.

5. Test Refreshes Manually: If you have a report that always seems to fail, try refreshing it manually in Power BI Service. Sometimes you can spot a dataset or step that’s consistently problematic.

6. Check All Data Sources: Make sure all data sources are actually up and reachable when the refresh kicks off (network drives, databases, cloud sources, etc.). Temporary network blips or source outages can cause this error.

7. Check for Antivirus/Firewall Issues: Occasionally, aggressive antivirus or firewall rules can interfere with gateway operations. Try temporarily disabling them (if you can) to rule this out.

8. Use a Gateway Cluster if Needed: If you’re seeing frequent failures under heavy load, setting up a gateway cluster can help distribute the refreshes and avoid bottlenecks.

9. Escalate with Details: If you’ve done all this and still see failures, gather logs, correlation IDs, and timestamps and reach out to Microsoft support. Even if you don’t have a paid support plan, posting this info here can sometimes get more attention.

If anyone’s spotted a specific pattern (certain times, only certain datasets, more frequent lately, etc.), definitely share it sometimes the smallest detail helps connect the dots. 

[ezaidi]

This issue is still persistent and has not gone away. We have tried the following:

1) updated gateway to latest version

2) created a secondary gateway cluster

3) I exported the logs but I cant really decipher the error

4) Reduced the frequency of scheduled refreshes on a number of reports and staggered schedules as much as possible.

5) We routinely refresh the reports manually when it fails. The reports are not the issue. This happens to big or small ones.

6) Unfortunately we do not have a support contract for fabric support plan.