Well, I have played with Splunk and the ELK (Elasticsearch - Logstash - Kibana) Stack for some time now; review these interesting Pages and Blogs:
both somehow, that was not sufficient. Recently i "stumpled" upon Grafana, a Graphic frontend for several Datasouces, including Elasticsearch (ES) from version 4.1.
There is a fine installation guide, but due to the fact I did not want to "pollude" my installation via the package system with "apt-get", i decided to get the tar file and unpack it into /data/grafana (see under Installing from binary tar file)
This is absolutly excellent and the server spun up and I could start using ES as a datasource at once.
The trouble began with I tried to set autostart up via the guide in http://docs.grafana.org/installation/debian/ - a lot of configurations was already saved under the manual startups adn workning, and several thing had to be corrected in the startup files to find these settings.
After a while I scratched it, and started with the generel method from https://gist.github.com/naholyr/4275302
Remember to add user:
My service.sh - called /etc/init.d/grafana-server is like this (notice the 2 times cd /data/grafana):
Make sure to change file ownership on all files to the grafana user:
And it works fine the cd to the /data/grafana ensures getting the right configs (As in the manual startup) and logs to /var/log/grafana/grafana.log