Durable Task Scheduler stopped dispatching work items to connected workers in UK South, twice, on five independent schedulers

Question

Five Durable Task Schedulers (Consumption SKU, UK South), each in its own subscription with its own task hub, stopped delivering orchestration and activity work items to their workers at the same times on 2026-10-29. Work resumed on all five without any action on our side, twice. Clients could still reach the schedulers throughout (scheduling new orchestrations, reading instance status and history), so new work was accepted and stayed Pending until dispatch resumed.

Each scheduler is used by one Azure Functions app (Flex Consumption, .NET 10 isolated worker) through the Durable Functions azureManaged storage provider. During both windows, every app's Durable worker logged this warning about every two minutes and did not recover until the service did:

No work items or health pings have been received from the server for an extended period of time. Attempting to reconnect.

No configuration, deployment or role change explains it. The windows match across five subscriptions to within a minute. The only restart in the period was of the dev app at 21:01:42, after work had already resumed on all five at about 21:01.