Forum Discussion
Advocate IGateway Stability Issues - v3000.338.2
Super UserHi david_busint,
Hi,
If the gateway stability issues started after upgrading to version 3000.338.2, I would recommend checking whether the issue is related to the new gateway release before making further configuration changes.
Please try these steps:
- Restart the On-premises data gateway service and check whether the issue persists.
- Open the gateway application → Diagnostics → Export logs, and review Windows Event Viewer for gateway service errors.
- Check CPU and memory usage during refreshes, and run the gateway Network ports test.
- If you're using a gateway cluster, ensure all gateway members are running the same version.
- If everything was working before the upgrade, test a previously stable gateway version in a controlled environment. Keep the gateway recovery key available before reinstalling.
My recommendation: If the issue started immediately after upgrading and the previous version works without problems, it may indicate a regression in version 3000.338.2. In that case, raise a Microsoft Support ticket with the gateway logs, error messages and timestamps.
Microsoft documentation:
https://learn.microsoft.com/en-us/data-integration/gateway/service-gateway-tshoot
Hope this helps!
Best Regards,
Maruthi
LinkedIn - http://www.linkedin.com/in/maruthi-siva-prasad/
X - Maruthi Siva Prasad - (@MaruthiSP) / X
david_busintHi maruthisp,
Thanks for the response.
- Done that, no change
- Going to see if I can get the logs
- CPU/Memory fine
- Not got a gateway cluster set up
- Tested and was stable on the May version, only have problems when moving to June+ version.
I'll see if I can see what's in the logs and if that doesn't tell us anything I will raise a ticket to MS.
Cheers,
David
- ShivekMaharaj
Community Champion
Hi david_busint,
Your latest testing narrows this down quite a bit.
The important change around June 2026 is that the on-premises data gateway started using the bundled Oracle Managed ODP.NET provider by default for Import workloads. Microsoft documents that change in the Oracle connector guidance, together with the MashupFlight_DisableOracleBundledOdacProviderV2 setting you have already applied to fall back to the previous provider.
Since the same gateway environment is stable on the May version but begins showing the online/offline flapping again on June and later builds, I would now compare the gateway logs around the exact timestamps where the status changes rather than changing more Oracle settings.
Microsoft's gateway troubleshooting guidance recommends exporting the gateway logs and checking Windows Event Viewer when the service becomes unstable. I would specifically look for whether PBIEgwService is restarting, crashing, losing its cloud connection, or only failing Oracle refresh processes while the service itself remains alive.
That distinction matters because an Oracle connector failure should not normally make the whole gateway repeatedly appear offline. If the gateway service itself is restarting or disconnecting, the issue is broader than an individual refresh.
I would therefore capture:
- gateway logs covering one offline/online cycle
- Windows Application and System events for the same timestamps
- the exact gateway build where the issue occurs
- whether PBIEgwService restarted or remained running
- one failed Oracle refresh Request ID and timestamp
Since Microsoft currently supports the October 2026 3000.338 release and your issue reproduces there while May remains stable, I think you have enough evidence for a support case if the logs show a service-level failure.I would include the fact that disabling the bundled Oracle provider restores connectivity but does not eliminate the gateway stability problem, because that helps separate the Oracle provider change from the gateway process instability itself.
AI-assisted drafting: AI was used to help structure and phrase this response. I reviewed and validated the technical content before posting.
