SignalR IIS Logs Flood With Logs

News Knowledge Base
, , , , , , , , , , , , ,
1

Resolution when SignalR - IIS Logs flood with logs .

Issue Description:
The Orchestrator server running out of disk space can potentially cause an outage. The IIS logs for Orchestrator are increasing substantially in size.

IIS logs contain a very large number of entries formatted like:

DATE TIME Orchestrator_IP_Address POST /signalr/hubs/signalr/poll clientProtocol=2.1&transport=longPolling&connectionData=[SOME_DATA] 443 - Robot_IP_Address SignalR.Client.NetStandard/2.4.1.0+(Microsoft+Windows+NT+10.0.14393.0) - 200 0 0 0


Resolution:

  1. Restart Robot Service on Robot_IP_Address and check if excessive log entries continue. Note: There could be more than one offending Robot_IP_Address, so check the IIS Log thoroughly. If the log entries continue from Robot_IP_Address(es), proceed to step 2.
  2. Adjust SignalR settings for Scalability in Orchestrator to be WebSocket only by going to Orchestrator -> Settings -> Scalability -> leave only WebSocket option checked -> Save.

image001.png

  1. Restart the Robot service after saving.



More details can be found in SignalR guide.
After upgrading the Orchestrator, revert the changes by checking all the options under the scalability settings within the Orchestrator.