ADF Managed VNet Integration Runtime Slow Processing/Timeouts in UK South

Question

I am experiencing slow processing and intermittent timeouts with the Managed VNet Integration Runtime in Azure Data Factory, specifically in the UK South region. The execution times are notably longer than usual and some activities are queued for a longer time leading to failure in pipeline.

• Issue region: UK South
• Integration runtime type: Managed VNet
• Effect: Pipelines are delayed, tasks queue for several minutes, and some jobs timeout or fail.
• An example of the output is below, where pipeline is failing even after running for longer hours than usual. We only usually see tasks queued for a few minutes (approx 15 mins) but now few of the tasks are sitting in queue more than 30 minutes leading to time-out, this has just started happening as of yesterday and nothing significant has changed at our end. Is there some sort of outage or service issue with ADF Integration Runtimes in UK South?
  

Answer 1

Hello !

Thank you for posting on Microsoft Learn Q&A.

I couldn't find any public Azure notice right now about an ADF or IR incident in UK South. I checked the Azure status site and don’t see a current regional advisory for Data Factory. Public pages don’t always catch narrow-scope capacity issues though your symptoms (activities stuck in Queued 30+ mins, then timing out) match regional capacity pressure on the Azure/Managed VNet IR.

You can bypass AutoResolve or busy shared capacity if any activities still run on AutoResolveIntegrationRuntime and create a dedicated Azure IR pinned to UK South and point the affected linked services to it.

At least you can avoid the shared pool that gets congested.

https://learn.microsoft.com/en-us/answers/questions/2281353/adf-uk-south-autoresolveintegrationruntime-excessi

Also try to split load across IRs Spin up a second Managed VNet IR in the same region and route some high-volume activities to it, so you can reduce queue contention at the IR level.

You can reach out to Azure support if your issue persists.