SwarmAgent stops working after photon caching

Hi, I’m using distributed rendering with swarm agent. My remote machine always stops after finishing photon caching and never works for the “Processing Mappings”.

18:53:26: Lightmass on PC-PC: 1:17 min total, 9.01 sec importing, 3.25 sec setup, 1:04 min photons, 2 ms processing, 0 ms extra exporting [0/38 mappings]. Threads: 87 ms total, 0 ms processing.
18:53:26: Lighting complete [Startup = 9.01 sec, Lighting = 1:08 min]
18:53:26: Failed to send back log file through Swarm!
18:53:26: [CloseConnection] Closing connection 1E98F7DB using handle 1E98F7DB
18:53:26: [CloseConnection] Connection confirmed for disconnection 1E98F7DB
18:53:26: [CloseConnection] Connection disconnected 1E98F7DB
18:53:26: [GetMessage] Safely returning to 1E98F7DB with no message
18:53:26: [MaintainConnections] Detected finished connection, closing (1E98F7D9)
18:53:26: [CloseConnection] Closing connection 1E98F7D9 using handle 1E98F7D9
18:53:26: [CloseConnection] Connection confirmed for disconnection 1E98F7D9
18:53:26: [CloseConnection] Closing bi-directional remote connection (1E98F7D9)
18:53:26: [CloseConnection] Closing connection 1E98F7D9 using handle 1E98F7D9
18:53:26: [CloseConnection] Connection already being disconnected, nothing to do 1E98F7D9
18:53:26: [CloseConnection] Connection disconnected 1E98F7D9
18:53:26: [MaintainConnections] Local connection has closed (1E98F7DB)
18:53:26: [MaintainConnections] Removed connection 1E98F7DB
18:53:26: [Maintain Jobs] Killing rogue Job "43E84291-4C018AA3-2F53638F-E7BB6503"
18:53:27: [MaintainConnections] Remote connection has closed (1E98F7D9)
18:53:27: [MaintainConnections] Removed connection 1E98F7D9
18:53:27: [MaintainConnections] All connections have closed

Any suggestions? Thank you!

Have you tried the steps from this answer?

Thank you .

I updated to 4.11.1 and part of the problem is solved. The remote agents can now work for the “mapping processing”. But they still stop before the entire baking is finished.

My experience from other distributed rendering solutions(e.g. Vray) is that all the render nodes should work together until the whole rendering job is done. But my swarm agents just stop working after a while and leave the rendering job to my local machine. The total rendering time is about 1600 seconds, the agents worked for only 200 seconds. There are enough space for cache. Is this normal for lightmass building?