This monitors the length of the logstash:beaver queue in redis, via the community logstash-queue-length plugin.
If this queue is growing, it implies logstash or one of its 'downstream outputs' is not processing log in a timely manner, or not at all.
Quite often, it's because ES has failed, or logstash process has died.
It can also mean though that your log processing is too much for the server to handle. Check resource graphs.