I had a gateway that hung over the weekend and about 20 reports failed to update.
The message in the service for Refresh History was "Timeout occurred when sending or receiving a request."
The details were:
Cluster URI: WABI-WEST-US-redirect.analysis.windows.net Activity ID: ccf178bc-c984-4433-a2d5-519274c8f31a Request ID: 21a5635a-4109-4847-a8fb-7ded1d4a490b Time: 2018-07-09 06:13:47Z
When I went to look at Settings, Manage Gateways, it said "You are good to go" but had a warning that one node was not responding.
I logged into that PC and launching the gateway app hung, so I rebooted. Then everything worked fine.
However, I had two other nodes that were operating just fine. Shouldn't the service have recognized that a node was timing out and tried another node? Isn't that the purpose of the cluster feature?