In the past few days I have come across the following error on two different instances of Logstash.
On the first instance I came across the mount point that elasticsearch was writing to became full. In this case I had to do the following to get everything back on it.
curl -XDELETE http://localhost:9200/_all
sudo service elasticsearch restart
Keep in mind the above will delete all data from elasticsearch. There may be a more graceful way of doing some cleanup but for this one it worked as I did not care about retaining any previous data.
On the second instance I saw this there was still plenty of available space on the mountpoint for elasticsearch so I just restarted elasticsearch and all was fine. Looking at the elasticsearch log file this is what it contained.