Hello,
I have the same problem here (We are running logstash 2.2).
I have a bunch of apache logs which I'm trying to get into elasticsearch.
My workflow is:
scp domain/access_log srvlogstash:/var/log/xxx
nxlog read on srvlogstash server the file and send to 5514 tcp the log, logstash send it to elasticsearch cluster.
I watch ndocs on elasticsearch, when the logs on access_log are indexed on elasticsearch cluster, I copy another access_log file
logstash got stucked about 2 days later I started the process, I had to restart it (and then restart nxlog too)
this machine is not on elasticsearch cluster as client, I'm not sure if it would work better if I add it as a elasticsearch client.
Regards,