How to recover from janusgraph exceptions?

[sinha-abhishek]

I need some help in how to recover from Janusgraph exceptions. I am running gremlin server on my server instance and making queries over websocket connection. The problem is whenever any query causes an exception like parallelMutate or sleep interrupted, the gremlin server seems to go into a bad state and starts throwing exceptions for all following gremlin queries made over websocket. The only option is to restart the server. I tried making new websocket connection for each query, however, it still does not respond to any further query once an exception has occurred unless I kill and start the service. Am I doing something wrong or does there need a particular way to recover from exceptions without killing the gremlin server.