Hello,
We installed a first ES node on a Debian server with the dedicated package distribution.
It works well. Environment variables are in /etc/default/elasticsearch.
Then, we have been tried to install a second ES node on this same Debian server.
It does not start correctly because it takes the configuration and data paths of the first node.
We have been tried many tests and workarounds but it does not work. We have copied the /etc/default/elasticsearch file to /etc/default/elasticsearch-masteronly and we have set an other path for ES_PATH_CONF variable.
The startup script takes in account the ES_PATH_CONF from /etc/default/elasticsearch. Thus, at startup, the node fails because the path is already locked by the other node (it is normal).
The problem occurs too when we try to start form the command line by passing the path configuration and the path data values, without systemd script.
It ignores ES_PATH_CONF variable 
https://www.elastic.co/guide/en/elasticsearch/reference/current/settings.html#config-files-location
Is the value hard-coded in the Debian package ?
On Debian, is it possible to install 2 nodes on the same server ?
Thanks for your help 