I worked through the codebase and that exception is ignorable. You can turn it off by setting org.stagemonitor.util.IOUtils to not log using your log configuration. That class is logging it, but the Elastic class calling it is deliberately ignoring it and doesn't want it logged - just can't control the stagemonitor logging.
The functionality is just ensuring nothing is left unconsumed in the stream, so is reading until timeout - and low traffic means it hits that regularly, but it's not a functionally useful exception, so best to just suppress reporting it